]> git.lizzy.rs Git - rust.git/log
rust.git
5 years agostabilize top level or-pats in if/while let.
Mazdak Farrokhzad [Fri, 11 Jan 2019 22:57:04 +0000 (23:57 +0100)]
stabilize top level or-pats in if/while let.

5 years agoFixing a typographical error.
David Sanders [Fri, 11 Jan 2019 21:58:13 +0000 (13:58 -0800)]
Fixing a typographical error.

5 years agoFix new hyperlinks in RELEASES.md
Alexander Ronald Altman [Fri, 11 Jan 2019 19:59:04 +0000 (13:59 -0600)]
Fix new hyperlinks in RELEASES.md

5 years agodon't unwrap unexpected tokens in `format!`
Andy Russell [Fri, 11 Jan 2019 17:40:05 +0000 (12:40 -0500)]
don't unwrap unexpected tokens in `format!`

Fixes #57512.

5 years agoCorrect RELEASES.md for 1.32.0
Alexander Ronald Altman [Fri, 11 Jan 2019 17:06:56 +0000 (11:06 -0600)]
Correct RELEASES.md for 1.32.0

The `into_to_from_bytes` feature was stabilized for `i128` and `u128` just like for the other integer types, but they seem to have been missed.

5 years agore-do docs for core::cmp
Steve Klabnik [Thu, 10 Jan 2019 20:21:01 +0000 (15:21 -0500)]
re-do docs for core::cmp

Fixes #32934

5 years agoAuto merge of #57470 - RalfJung:miri, r=oli-obk
bors [Fri, 11 Jan 2019 16:28:45 +0000 (16:28 +0000)]
Auto merge of #57470 - RalfJung:miri, r=oli-obk

update miri

r? @oli-obk

5 years agoAuto merge of #57355 - arielb1:correct-subst, r=nikomatsakis
bors [Fri, 11 Jan 2019 12:04:24 +0000 (12:04 +0000)]
Auto merge of #57355 - arielb1:correct-subst, r=nikomatsakis

use the correct supertrait substitution in `object_ty_for_trait`

beta-nominating because regression.

Fixes #57156.

5 years agoRemove unneeded but benign change
Oliver Scherer [Fri, 11 Jan 2019 11:19:08 +0000 (12:19 +0100)]
Remove unneeded but benign change

5 years agoFix undefined behavior
Jethro Beekman [Fri, 11 Jan 2019 09:30:08 +0000 (15:00 +0530)]
Fix undefined behavior

5 years agoAdd a profiles section to the manifest
Nick Cameron [Fri, 11 Jan 2019 05:13:45 +0000 (18:13 +1300)]
Add a profiles section to the manifest

5 years agoAddress comments
John Kåre Alsaker [Fri, 11 Jan 2019 03:58:46 +0000 (04:58 +0100)]
Address comments

5 years agoUpdate tests
John Kåre Alsaker [Tue, 1 Jan 2019 18:08:25 +0000 (19:08 +0100)]
Update tests

5 years agoMake more passes incremental
John Kåre Alsaker [Wed, 6 Jun 2018 20:13:52 +0000 (22:13 +0200)]
Make more passes incremental

5 years agoadd test for pub extern crate
DebugSteven [Fri, 11 Jan 2019 02:18:46 +0000 (21:18 -0500)]
add test for pub extern crate

5 years agoinline pub extern crate statements
DebugSteven [Fri, 11 Jan 2019 01:27:44 +0000 (20:27 -0500)]
inline pub extern crate statements

5 years agoRemove submodule step from README
Nick Cameron [Fri, 11 Jan 2019 01:27:12 +0000 (14:27 +1300)]
Remove submodule step from README

Since the bootstrap does it now

5 years agoAuto merge of #57471 - Aaronepower:master, r=Aaronepower
bors [Fri, 11 Jan 2019 00:20:04 +0000 (00:20 +0000)]
Auto merge of #57471 - Aaronepower:master, r=Aaronepower

Updated RELEASES.md for 1.32.0

[Rendered](https://github.com/Aaronepower/rust/blob/master/RELEASES.md)

r? @Mark-Simulacrum
cc @rust-lang/release

5 years agoUpdate RELEASES.md
Aaron Power [Thu, 10 Jan 2019 22:28:18 +0000 (23:28 +0100)]
Update RELEASES.md

5 years agoUpdate src/libstd/path.rs
Mazdak Farrokhzad [Thu, 10 Jan 2019 22:08:42 +0000 (17:08 -0500)]
Update src/libstd/path.rs

Co-Authored-By: steveklabnik <steve@steveklabnik.com>
5 years agonote that FromStr does not work for borrowed types
Steve Klabnik [Thu, 10 Jan 2019 20:40:05 +0000 (15:40 -0500)]
note that FromStr does not work for borrowed types

Fixes #47757

5 years agoUpdate RELEASES.md
Aaron Power [Thu, 10 Jan 2019 22:07:38 +0000 (23:07 +0100)]
Update RELEASES.md

5 years agoAuto merge of #57484 - alexcrichton:fix-nightlies, r=Mark-Simulacru
bors [Thu, 10 Jan 2019 21:31:34 +0000 (21:31 +0000)]
Auto merge of #57484 - alexcrichton:fix-nightlies, r=Mark-Simulacru

Integrate miri into build-manifest

This fixes a mistake where miri was accidentally left out of the
build-manifest parsing, meaning that today's nightly generated a
manifest with invalid urls!

Fixes #57488.

5 years agomake note of one more normalization that Paths do
Steve Klabnik [Thu, 10 Jan 2019 20:30:36 +0000 (15:30 -0500)]
make note of one more normalization that Paths do

Fixes #29008

5 years agoAdd a fast path for identical regions in lub_concrete_regions
Björn Steinbrink [Thu, 10 Jan 2019 18:28:42 +0000 (19:28 +0100)]
Add a fast path for identical regions in lub_concrete_regions

In functions with lots of region constraint, if the fixed point
iteration converges only slowly, a lot of the var/var constraints will
have equal regions most of the time. Yet, we still perform the LUB
calculation and try to intern the result. Especially the latter incurs
quite some overhead.

This reduces the take taken by the item bodies checking pass for the
unicode_normalization crate by about 75%.

5 years agoDrop "solved" constraints during region expansion
Björn Steinbrink [Wed, 9 Jan 2019 17:40:30 +0000 (18:40 +0100)]
Drop "solved" constraints during region expansion

Once a region has been expanded to cover a fixed region, a corresponding
RegSubVar constraint won't have any effect on the expansion anymore, the
same is true for constraints where the variable on the RHS has already
reached static scope. By removing those constraints from the set that
we're iterating over, we remove a lot of needless overhead in case of
slow convergences (i.e. lots of iterations).

For the unicode_normalization crate, this about cuts the time required
for item_bodies checking in half.

5 years agouse structured suggestion when casting a reference
Andy Russell [Thu, 10 Jan 2019 18:42:59 +0000 (13:42 -0500)]
use structured suggestion when casting a reference

5 years agoIntegrate miri into build-manifest
Alex Crichton [Thu, 10 Jan 2019 01:14:11 +0000 (17:14 -0800)]
Integrate miri into build-manifest

This fixes a mistake where miri was accidentally left out of the
build-manifest parsing, meaning that today's nightly generated a
manifest with invalid urls!

5 years agostd: Render large exit codes as hex on Windows
Alex Crichton [Wed, 9 Jan 2019 17:35:09 +0000 (09:35 -0800)]
std: Render large exit codes as hex on Windows

On Windows process exit codes are never signals but rather always 32-bit
integers. Most faults like segfaults and such end up having large
integers used to represent them, like STATUS_ACCESS_VIOLATION being
0xC0000005. Currently, however, when an `ExitStatus` is printed this
ends up getting rendered as 3221225477 which is somewhat more difficult
to debug.

This commit adds a branch in `Display for ExitStatus` on Windows which
handles exit statuses where the high bit is set and prints those exit
statuses as hex instead of with decimals. This will hopefully preserve
the current display for small exit statuses (like `exit code: 22`), but
assist in quickly debugging segfaults/access violations/etc. I've
found at least that the hex codes are easier to search for than decimal.

I wasn't able to find any official documentation saying that all system
exit codes have the high bit set, but I figure it's a good enough
heuristic for now.

5 years agoUpdate Cargo.lock
Hugues de Valon [Thu, 10 Jan 2019 12:42:14 +0000 (12:42 +0000)]
Update Cargo.lock

5 years agoApply suggestions from code review
Aaron Power [Thu, 10 Jan 2019 09:08:32 +0000 (10:08 +0100)]
Apply suggestions from code review

Co-Authored-By: Aaronepower <Aaronepower@users.noreply.github.com>
5 years agoUpdate RELEASES.md
Aaron Power [Thu, 10 Jan 2019 09:08:07 +0000 (10:08 +0100)]
Update RELEASES.md

5 years agoactually take a slice in this example
Steve Klabnik [Tue, 8 Jan 2019 19:58:29 +0000 (14:58 -0500)]
actually take a slice in this example

Fixes #45678

5 years agoclarify resolve typo suggestion
Andy Russell [Wed, 9 Jan 2019 19:11:00 +0000 (14:11 -0500)]
clarify resolve typo suggestion

Include the kind of the binding that we're suggesting, and use a
structured suggestion.

5 years agosave-analysis: Get path def from parent in case there's no def for the path itself.
Emilio Cobos Álvarez [Wed, 9 Jan 2019 16:31:16 +0000 (17:31 +0100)]
save-analysis: Get path def from parent in case there's no def for the path itself.

This fixes #57462.

The relevant part from the hir type collector is:

```
DEBUG 2019-01-09T15:42:58Z: rustc::hir::map::collector: hir_map: NodeId(32) => Entry { parent: NodeId(33), dep_node: 4294967040, node: Expr(expr(32: <Foo>::new)) }
DEBUG 2019-01-09T15:42:58Z: rustc::hir::map::collector: hir_map: NodeId(48) => Entry { parent: NodeId(32), dep_node: 4294967040, node: Ty(type(Foo)) }
DEBUG 2019-01-09T15:42:58Z: rustc::hir::map::collector: hir_map: NodeId(30) => Entry { parent: NodeId(48), dep_node: 4294967040, node: PathSegment(PathSegment { ident: Foo#0, id: Some(NodeId(30)), def: Some(Err), args: None, infer_types: true }) }
DEBUG 2019-01-09T15:42:58Z: rustc::hir::map::collector: hir_map: NodeId(31) => Entry { parent: NodeId(32), dep_node: 4294967040, node: PathSegment(PathSegment { ident: new#0, id: Some(NodeId(31)), def: Some(Err), args: None, infer_types: true }) }
```

We have the right ID when looking for NodeId(31) and try with NodeId(32) (which
is the right thing to look for) from get_path_data, but not for the segments
that we write from `write_sub_paths_truncated`.

Basically `process_path` takes an id which is always the parent, and that we
fall back to in `get_path_data()`, so we get the right result for the last path
segment, but not for the other segments that get written to from
`write_sub_paths_truncated`.

I think we can stop passing the explicit id around to `get_path_data` now, will
consider sending that as a followup.

5 years agoNot seeing the forest because there are too many trees in the way
Oliver Scherer [Wed, 9 Jan 2019 15:34:31 +0000 (16:34 +0100)]
Not seeing the forest because there are too many trees in the way

5 years agoExplain the arguments of the `mk_*_eval_cx` functions
Oliver Scherer [Wed, 9 Jan 2019 15:27:33 +0000 (16:27 +0100)]
Explain the arguments of the `mk_*_eval_cx` functions

5 years agoUpdated RELEASES.md for 1.32.0
Aaron Power [Wed, 9 Jan 2019 15:12:12 +0000 (16:12 +0100)]
Updated RELEASES.md for 1.32.0

5 years agoDocument the `mk_*_eval_cx` functions
Oliver Scherer [Wed, 9 Jan 2019 15:08:07 +0000 (16:08 +0100)]
Document the `mk_*_eval_cx` functions

5 years agoupdate miri
Ralf Jung [Wed, 9 Jan 2019 15:03:01 +0000 (16:03 +0100)]
update miri

5 years agoUpdate cc crate to include Armv8-M changes
Hugues de Valon [Wed, 9 Jan 2019 14:51:22 +0000 (14:51 +0000)]
Update cc crate to include Armv8-M changes

5 years agoReplace Arm GCC to a more recent version
Hugues de Valon [Thu, 3 Jan 2019 13:42:02 +0000 (13:42 +0000)]
Replace Arm GCC to a more recent version

The version included in Ubuntu 16.04 repositories in the dist-various-1
docker, Arm GCC version 4.9, does not support the new Armv8-M
architecture.
This commit adds the team-gcc-arm-embedded PPA to get through APT a
newer version of Arm GCC.

5 years agoAdd dist builder for Armv8-M Mainline
Hugues de Valon [Tue, 18 Dec 2018 16:30:34 +0000 (16:30 +0000)]
Add dist builder for Armv8-M Mainline

This commit adds the Armv8-M Mainline target in the list of targets that
get their dist components built. It also update the build-manifest so
that this target gets also its dist components uploaded.

5 years agoAuto merge of #57419 - cramertj:pin-set, r=withouboats
bors [Wed, 9 Jan 2019 13:48:37 +0000 (13:48 +0000)]
Auto merge of #57419 - cramertj:pin-set, r=withouboats

Reborrow Pin<P> using &mut in `Pin::set`

Fixes https://github.com/rust-lang/rust/issues/57339.

This makes it possible to call `.set` multiple times without
using `.as_mut()` first to reborrow the pointer.

r? @withoutboats
cc @rust-lang/libs

5 years agoMetadataOnlyCodegenBackend: remove `is_inline` call
lqd [Wed, 9 Jan 2019 13:21:46 +0000 (14:21 +0100)]
MetadataOnlyCodegenBackend: remove `is_inline` call

This function has no side effects, and its result is ignored.

5 years agoExposing enclave image-base to the enclave application
Vardhan Thigle [Wed, 9 Jan 2019 11:50:37 +0000 (17:20 +0530)]
Exposing enclave image-base to the enclave application

image-base could be used by crates like backtrace to providing to make
symbol resolution easier.

5 years agoRemove outdated comment
king6cong [Wed, 9 Jan 2019 11:42:25 +0000 (19:42 +0800)]
Remove outdated comment

5 years agoAuto merge of #56614 - Zoxc:query-perf2, r=michaelwoerister
bors [Wed, 9 Jan 2019 11:08:14 +0000 (11:08 +0000)]
Auto merge of #56614 - Zoxc:query-perf2, r=michaelwoerister

Replace LockCell with atomic types

Split from https://github.com/rust-lang/rust/pull/56509

r? @michaelwoerister

5 years agoFix irrefutable slice patterns in const fn
Oliver Scherer [Wed, 9 Jan 2019 10:32:56 +0000 (11:32 +0100)]
Fix irrefutable slice patterns in const fn

5 years agoconst fn feature gate is not needed anymore in a lot of tests
Oliver Scherer [Sun, 30 Dec 2018 18:20:53 +0000 (19:20 +0100)]
const fn feature gate is not needed anymore in a lot of tests

5 years agoClarify const_let comment
Oliver Scherer [Sun, 30 Dec 2018 17:57:31 +0000 (18:57 +0100)]
Clarify const_let comment

5 years agoStabilize `let` bindings and destructuring in constants and const fn
Oliver Scherer [Fri, 28 Dec 2018 19:05:22 +0000 (20:05 +0100)]
Stabilize `let` bindings and destructuring in constants and const fn

5 years agoAuto merge of #57086 - oli-obk:miri_dist, r=kennytm
bors [Wed, 9 Jan 2019 06:42:11 +0000 (06:42 +0000)]
Auto merge of #57086 - oli-obk:miri_dist, r=kennytm

Prepare everything for distributing miri via rustup

The next step is to tell rustup about `cargo-miri` in https://github.com/rust-lang/rustup.rs/blob/31935e5f633a5acd3a203d23b61d1556c64a821c/src/rustup/lib.rs#L28 and https://github.com/rust-lang/rustup.rs/blob/1ccd706d1d572c777c5134bd7db7aa1a8df7f278/src/rustup-win-installer/src/lib.rs#L29

5 years agodocs: Fix some 'second-edition' links
Philipp Hansch [Wed, 9 Jan 2019 06:15:52 +0000 (07:15 +0100)]
docs: Fix some 'second-edition' links

5 years agoChange `String` to `&'static str` in `ParseResult::Failure`.
Nicholas Nethercote [Wed, 9 Jan 2019 04:16:19 +0000 (15:16 +1100)]
Change `String` to `&'static str` in `ParseResult::Failure`.

This avoids 770,000 allocations when compiling the `html5ever`
benchmark, reducing instruction counts by up to 2%.

5 years agoMerge pull request #1 from Centril/redo-vec-set_len-docs-adjust
scottmcm [Wed, 9 Jan 2019 03:41:32 +0000 (19:41 -0800)]
Merge pull request #1 from Centril/redo-vec-set_len-docs-adjust

Explain safety for `vec.set_len(0)`

5 years agoaddressing Niko's comments
Blitzerr [Wed, 9 Jan 2019 03:13:50 +0000 (19:13 -0800)]
addressing Niko's comments

5 years agoexplain safety for vec.set_len(0)
Mazdak Farrokhzad [Wed, 9 Jan 2019 03:17:24 +0000 (04:17 +0100)]
explain safety for vec.set_len(0)

5 years ago[Cleanup] This is the first in the series of removals of with_freevars usage.
Blitzerr [Sat, 22 Dec 2018 23:06:14 +0000 (15:06 -0800)]
[Cleanup] This is the first in the series of removals of with_freevars usage.

Currently, there are many places in rustc, where we use
with_freevars to iterate over freevars of a closure. The
problem with this is the argument to with_freevars is a
NodeId and this will get in the way of our eventual goal
of solving for issue (#53488), sub-issue (#56905)

5 years agoSome more refactoring.
Blitzerr [Sat, 22 Dec 2018 04:00:11 +0000 (20:00 -0800)]
Some more refactoring.

Change the key of UpvarListMap from DefId to ast::NodeId

5 years agoCreating the vector using with_capacity to avoid re-allocation later on (#56905)
Blitzerr [Wed, 19 Dec 2018 04:54:35 +0000 (20:54 -0800)]
Creating the vector using with_capacity to avoid re-allocation later on (#56905)

5 years agoIssue 56905
Blitzerr [Mon, 17 Dec 2018 01:01:57 +0000 (17:01 -0800)]
Issue 56905

Adding a map to TypeckTables to get the list of all the Upvars
given a closureID. This is help us get rid of the recurring
pattern in the codebase of iterating over the free vars
using with_freevars.

5 years agoConsolidate equality constraints error message
varkor [Tue, 8 Jan 2019 23:53:43 +0000 (23:53 +0000)]
Consolidate equality constraints error message

5 years agoAdd issue reference to E0202 message
varkor [Tue, 8 Jan 2019 23:50:51 +0000 (23:50 +0000)]
Add issue reference to E0202 message

5 years agoChange std::error::Error trait documentation to talk about `source` instead of `cause`
Czipperz [Sat, 29 Dec 2018 07:09:21 +0000 (02:09 -0500)]
Change std::error::Error trait documentation to talk about `source` instead of `cause`

5 years agoAuto merge of #56407 - GuillaumeGomez:missing-docs-reexported-macros, r=varkor
bors [Tue, 8 Jan 2019 22:16:13 +0000 (22:16 +0000)]
Auto merge of #56407 - GuillaumeGomez:missing-docs-reexported-macros, r=varkor

check missing docs for reexported macros as well

Fixes #56334.

5 years agoMisc cleanups
Shotaro Yamada [Sun, 6 Jan 2019 06:36:14 +0000 (15:36 +0900)]
Misc cleanups

5 years agolldb_batchmode.py: try `import _thread` for Python 3
Josh Stone [Tue, 8 Jan 2019 21:19:50 +0000 (13:19 -0800)]
lldb_batchmode.py: try `import _thread` for Python 3

5 years agoAuto merge of #56638 - matthewjasper:remove-ref-region, r=nikomatsakis
bors [Tue, 8 Jan 2019 19:33:20 +0000 (19:33 +0000)]
Auto merge of #56638 - matthewjasper:remove-ref-region, r=nikomatsakis

Remove some `Region`s from HAIR

Use `ReErased` for any regions that need to be created in RValue::Ref
in MIR generation. We will change them to all to `ReVar` before borrow
checking anyway.

r? @nikomatsakis

5 years agoUpdate stdsimd submodule
Guillaume Gomez [Mon, 7 Jan 2019 15:20:25 +0000 (16:20 +0100)]
Update stdsimd submodule

5 years agoimprove non_camel_case_types diagnostics
Andy Russell [Fri, 4 Jan 2019 20:59:07 +0000 (15:59 -0500)]
improve non_camel_case_types diagnostics

Use a structured suggestion and tighten the span to just the identifier.

5 years agoimprove non_upper_case_globals diagnostics
Andy Russell [Fri, 4 Jan 2019 20:00:15 +0000 (15:00 -0500)]
improve non_upper_case_globals diagnostics

Use a structured suggestion and tighten the span to just the identifier.

5 years agoimprove non_snake_case diagnostics
Andy Russell [Fri, 4 Jan 2019 15:19:52 +0000 (10:19 -0500)]
improve non_snake_case diagnostics

Use a structured suggestion and tighten the span to just the identifier.

5 years agoMove diagnostics out from QueryJob and optimize for the case with no diagnostics
John Kåre Alsaker [Fri, 7 Dec 2018 02:04:23 +0000 (03:04 +0100)]
Move diagnostics out from QueryJob and optimize for the case with no diagnostics

5 years agoAuto merge of #57429 - alexcrichton:fix-dist, r=Mark-Simulacrum
bors [Tue, 8 Jan 2019 16:47:27 +0000 (16:47 +0000)]
Auto merge of #57429 - alexcrichton:fix-dist, r=Mark-Simulacrum

Build LLVM with -static-libstdc++ on dist builds

This commit is intended on fixing a regression from #57286 where the
distributed LLVM shared library unfortunately depends on a dynamic copy
of libstdc++, meaning we're no longer as binary compatible as we
thought! This tweaks the build of LLVM as out distribution is slightly
different now, and is hoped to fix the issue.

Closes #57426

5 years agoSupporting backtrace for x86_64-fortanix-unknown-sgx.
Vardhan Thigle [Thu, 3 Jan 2019 05:16:22 +0000 (10:46 +0530)]
Supporting backtrace for x86_64-fortanix-unknown-sgx.

5 years agoMore feature whitelisting of winapi
Oliver Scherer [Mon, 7 Jan 2019 16:17:41 +0000 (17:17 +0100)]
More feature whitelisting of winapi

5 years agoAdjust cargo workspace hack to miri/cargo interactions
Oliver Scherer [Tue, 1 Jan 2019 13:36:05 +0000 (14:36 +0100)]
Adjust cargo workspace hack to miri/cargo interactions

5 years agofixup
Oliver Scherer [Sun, 23 Dec 2018 20:31:45 +0000 (21:31 +0100)]
fixup

5 years agoPrepare everything for distributing miri via rustup
Oliver Scherer [Sun, 23 Dec 2018 20:20:35 +0000 (21:20 +0100)]
Prepare everything for distributing miri via rustup

5 years agoIgnore some IDE-local files
Oliver Scherer [Sun, 23 Dec 2018 20:21:02 +0000 (21:21 +0100)]
Ignore some IDE-local files

5 years agoAuto merge of #57114 - Zoxc:query-perf11, r=michaelwoerister
bors [Tue, 8 Jan 2019 14:11:19 +0000 (14:11 +0000)]
Auto merge of #57114 - Zoxc:query-perf11, r=michaelwoerister

Clean up and optimize OpenTask / read_index

r? @michaelwoerister

5 years agoAuto merge of #56988 - alexcrichton:monotonic-instant, r=sfackler
bors [Tue, 8 Jan 2019 11:30:19 +0000 (11:30 +0000)]
Auto merge of #56988 - alexcrichton:monotonic-instant, r=sfackler

std: Force `Instant::now()` to be monotonic

This commit is an attempt to force `Instant::now` to be monotonic
through any means possible. We tried relying on OS/hardware/clock
implementations, but those seem buggy enough that we can't rely on them
in practice. This commit implements the same hammer Firefox recently
implemented (noted in #56612) which is to just keep whatever the lastest
`Instant::now()` return value was in memory, returning that instead of
the OS looks like it's moving backwards.

Closes #48514
Closes #49281
cc #51648
cc #56560
Closes #56612
Closes #56940

5 years agoMake `mk_eval_cx` private to const eval
Oliver Scherer [Tue, 8 Jan 2019 09:54:40 +0000 (10:54 +0100)]
Make `mk_eval_cx` private to const eval

5 years agoManually push a stack frame where no valid frame is needed
Oliver Scherer [Tue, 8 Jan 2019 09:54:24 +0000 (10:54 +0100)]
Manually push a stack frame where no valid frame is needed

5 years agoAuto merge of #57332 - Dylan-DPC:feature/stabilise-cfg-attr, r=Centril
bors [Tue, 8 Jan 2019 07:49:13 +0000 (07:49 +0000)]
Auto merge of #57332 - Dylan-DPC:feature/stabilise-cfg-attr, r=Centril

stabilize cfg_attr_multi

Stabilizes cfg_attr_multi feature

Related to #54881

Will add the lint in a seperate PR

r? @Centril

5 years agoRemove `CrateNum::Invalid`.
Nicholas Nethercote [Tue, 8 Jan 2019 05:13:22 +0000 (16:13 +1100)]
Remove `CrateNum::Invalid`.

It's unused.

5 years agoAdd link destination for `read-ownership`
Dylan MacKenzie [Tue, 8 Jan 2019 05:03:41 +0000 (21:03 -0800)]
Add link destination for `read-ownership`

5 years agoMake `TokenStream` less recursive.
Nicholas Nethercote [Wed, 19 Dec 2018 03:53:52 +0000 (14:53 +1100)]
Make `TokenStream` less recursive.

`TokenStream` is currently recursive in *two* ways:

- the `TokenTree` variant contains a `ThinTokenStream`, which can
  contain a `TokenStream`;

- the `TokenStream` variant contains a `Vec<TokenStream>`.

The latter is not necessary and causes significant complexity. This
commit replaces it with the simpler `Vec<(TokenTree, IsJoint)>`.

This reduces complexity significantly. In particular, `StreamCursor` is
eliminated, and `Cursor` becomes much simpler, consisting now of just a
`TokenStream` and an index.

The commit also removes the `Extend` impl for `TokenStream`, because it
is only used in tests. (The commit also removes those tests.)

Overall, the commit reduces the number of lines of code by almost 200.

5 years agoBuild LLVM with -static-libstdc++ on dist builds
Alex Crichton [Mon, 7 Jan 2019 21:47:35 +0000 (13:47 -0800)]
Build LLVM with -static-libstdc++ on dist builds

This commit is intended on fixing a regression from #57286 where the
distributed LLVM shared library unfortunately depends on a dynamic copy
of libstdc++, meaning we're no longer as binary compatible as we
thought! This tweaks the build of LLVM as out distribution is slightly
different now, and is hoped to fix the issue.

Closes #57426

5 years agoAuto merge of #57095 - Zoxc:prof-fix, r=michaelwoerister
bors [Tue, 8 Jan 2019 02:00:12 +0000 (02:00 +0000)]
Auto merge of #57095 - Zoxc:prof-fix, r=michaelwoerister

Fix and optimize query profiling

r? @michaelwoerister

cc @wesleywiser

5 years agosave-analysis: use a fallback when access levels couldn't be computed
Igor Matuszewski [Mon, 7 Jan 2019 22:55:55 +0000 (23:55 +0100)]
save-analysis: use a fallback when access levels couldn't be computed

5 years agoReborrow Pin<P> using &mut in `Pin::set`
Taylor Cramer [Mon, 7 Jan 2019 19:45:34 +0000 (11:45 -0800)]
Reborrow Pin<P> using &mut in `Pin::set`

This makes it possible to call `.set` multiple times without
using `.as_mut()` first to reborrow the pointer.

5 years agoAuto merge of #57303 - matthiaskrgr:clippy_submodule_upd, r=oli-obk
bors [Mon, 7 Jan 2019 19:44:00 +0000 (19:44 +0000)]
Auto merge of #57303 - matthiaskrgr:clippy_submodule_upd, r=oli-obk

submodules: update clippy and rls

Fixes clippy toolstate

Changes:
````
Update to latest compiletest-rs release
add testcase for #3462
deps: bump rustc_tools_util version from 0.1.0 to 0.1.1 just in case...
rustc_tool_utils: fix failure to create proper non-repo version string when used in crates on crates.io, bump version
UI test cleanup: Extract ifs_same_cond tests
UI test cleanup: Extract for_kv_map lint tests
Fix test for rust-lang/rust#57250
Limit infinite_iter collect() check to known types
Some improvements to util documentation
Use hashset for name blacklist
Reformat random_state tests
Use node_id_to_type_opt instead of node_it_to_type in random_state
Check pattern equality while checking declaration equality
random_state lint
Use an FxHashSet for valid idents in documentation lint
Fix suggestion for unnecessary_ref lint
Update CONTRIBUTING.md for rustfix tests
Update .fixed files via update-references.sh
Run rustfix on first UI test
Use WIP branch for compiletest_rs
````

Also updates RLS and removes the patching of rustc_tool_utils from cargo.toml

RLS changes:
````
update clippy hash and rustc_tools_util and use rustc_tools_util from crates.io
Work around https://github.com/rust-lang/rust/pull/55937
Update Clippy... again
Update Clippy
Update clippy
````
r? @oli-obk

5 years agobless you
dylan_DPC [Sun, 6 Jan 2019 12:07:28 +0000 (17:37 +0530)]
bless you

5 years agoremove unwanted stage0 line, fix style
dylan_DPC [Fri, 4 Jan 2019 17:33:36 +0000 (23:03 +0530)]
remove unwanted stage0 line, fix style

5 years agoremove unused imports and feature gate from tests
dylan_DPC [Fri, 4 Jan 2019 11:51:29 +0000 (17:21 +0530)]
remove unused imports and feature gate from tests

5 years agoCodegen: run the collector only once
lqd [Mon, 7 Jan 2019 18:48:45 +0000 (19:48 +0100)]
Codegen: run the collector only once

Use the `collect_and_partition_mono_items ` query to avoid calling the collector

5 years agostabilise cfg_attr
dylan_DPC [Fri, 4 Jan 2019 07:07:27 +0000 (12:37 +0530)]
stabilise cfg_attr

5 years agoRevert "Auto merge of #57101 - o01eg:fix-57014, r=alexcrichton"
Matthias Krüger [Sun, 6 Jan 2019 12:55:15 +0000 (13:55 +0100)]
Revert "Auto merge of #57101 - o01eg:fix-57014, r=alexcrichton"

This reverts commit 68614265d312fc2cbe8a696f7dabb9416eb6f221, reversing
changes made to cae623c5ce12df8f237264d8f2c31fdaa664c382.

Should fix tools on windows.

Reopens #57014