Commit Graph

123375 Commits

Author SHA1 Message Date
Manish Goregaokar
dceafac96d
Rollup merge of #74347 - jyn514:ive-got-a-small-query-for-you, r=eddyb
Initialize default providers only once

This avoids copying a new `Providers` struct for each downstream crate
that wants to use it.

Follow-up to https://github.com/rust-lang/rust/pull/74283 without the perf hit.

r? @eddyb
2020-07-15 11:01:29 -07:00
Manish Goregaokar
bee28990d3
Rollup merge of #74344 - estebank:stringly-wobbly, r=eddyb
Remove string comparison and use diagnostic item instead

r? @eddyb
2020-07-15 11:01:28 -07:00
Manish Goregaokar
f4bbd0e607
Rollup merge of #74337 - estebank:ty-parse-recovery, r=varkor
Handle case of incomplete local ty more gracefully

When encountering a local binding with a type that isn't completed, the
parser will reach a `=` token. When this happen, consider the type
"complete" as far as the parser is concerned to avoid further errors
being emitted by parse recovery logic.
2020-07-15 11:01:26 -07:00
Manish Goregaokar
64a27880de
Rollup merge of #74294 - msirringhaus:master, r=pietroalbini
Update cross-compilation README

README seemed rather out of date. I hope the information in my PR is now correct (it was more or less assembled by asking in zulip and learning-by-doing).
2020-07-15 11:01:24 -07:00
Manish Goregaokar
b7cf2cbd78
Rollup merge of #74291 - regexident:from-docs, r=GuillaumeGomez
Added docs for `From<c_int>` for `ExitStatus`

Partially addresses https://github.com/rust-lang/rust/issues/51430
2020-07-15 11:01:22 -07:00
Manish Goregaokar
47f6a9d797
Rollup merge of #74276 - lcnr:discriminant-kind-what, r=nagisa
improve DiscriminantKind handling

Adds a lang item `discriminant_type` for the associated type `DiscriminantKind::Discriminant`.

Changes the discriminant of generators from `i32` to `u32`, which should not be observable to fix an
oversight where MIR was using `u32` and codegen and typeck used `i32`.
2020-07-15 11:01:20 -07:00
Manish Goregaokar
0d07db98ab
Rollup merge of #74218 - GuillaumeGomez:search-results-bottom-margin, r=Dylan-DPC
Add margin after doc search results

I found it not really on computer that the last result is right at the bottom of the page. I find it better with margin below (especially when you hover the last element!). A screenshot to show the result:

![Screenshot from 2020-07-10 16-32-23](https://user-images.githubusercontent.com/3050060/87166097-6103a580-c2cb-11ea-81a8-12772cf20f64.png)

r? @kinnison
cc @rust-lang/rustdoc @Manishearth @jyn514
2020-07-15 11:01:18 -07:00
Manish Goregaokar
efad203144
Rollup merge of #74196 - GuillaumeGomez:auto-collapse-implementors, r=Manishearth
Add option to collapse automatically implementors

Fixes #73403

It adds an option (enabled by default) which collapses all implementors impl blocks.

r? @kinnison
cc @rust-lang/rustdoc
2020-07-15 11:01:16 -07:00
Manish Goregaokar
0bb16c8efd
Rollup merge of #74119 - nnethercote:rm-Compiler-compile, r=Mark-Simulacrum
Remove `Compiler::compile()`.

It's unused.

r? @Mark-Simulacrum
2020-07-15 11:01:13 -07:00
Manish Goregaokar
1527126e07
Rollup merge of #73959 - GuillaumeGomez:cleanup-e0716, r=Dylan-DPC
Clean up E0716 explanation

r? @Dylan-DPC
2020-07-15 11:01:11 -07:00
Manish Goregaokar
7307af1f1c
Rollup merge of #73918 - GuillaumeGomez:cleanup-e0715, r=Dylan-DPC
Clean up E0715 explanation

r? @Dylan-DPC
2020-07-15 11:01:08 -07:00
Manish Goregaokar
af3d4cb936
Rollup merge of #72973 - msizanoen1:riscv-host, r=pietroalbini
RISC-V GNU/Linux as host platform

This PR add a new builder named `dist-riscv64-linux` that builds the compiler toolchain for RISC-V 64-bit GNU/Linux.

r? @alexcrichton
2020-07-15 11:01:02 -07:00
Guillaume Gomez
39d99ea6e6 Improve settings wording 2020-07-15 17:40:20 +02:00
Joshua Nelson
f6764c42ab Initialize default providers only once
This avoids copying a new `Providers` struct for each downstream crate
that wants to use it.
2020-07-15 10:53:36 -04:00
bors
7e11379f3b Auto merge of #74113 - lcnr:type-dependent-consts-2, r=eddyb
Support const args in type dependent paths (Take 2)

once more, except it is sound this time 🥰 previously #71154

-----
```rust
#![feature(const_generics)]

struct A;
impl A {
    fn foo<const N: usize>(&self) -> usize { N }
}
struct B;
impl B {
    fn foo<const N: usize>(&self) -> usize { 42 }
}

fn main() {
    let a = A;
    a.foo::<7>();
}
```
When calling `type_of` for generic const arguments, we now use the `TypeckTables` of the surrounding body to get the expected type.

This alone causes cycle errors though, as we now have `typeck_tables_of(main)` -> `...` ->
`type_of(main_ANON0 := 7)` -> `typeck_tables_of(main)`  (see https://github.com/rust-lang/rust/issues/68400#issuecomment-611760290)

To prevent this we must not call `type_of(const_arg)` during `typeck_tables_of`. This is achieved by
calling `type_of(param_def_id)` instead.

We have to somehow remember the `DefId` of the param through all of typeck, which is done using the
struct `ty::WithOptConstParam<DefId>`, which replaces `DefId` where needed and contains an `Option<DefId>` to
be able to store the const parameter in case it exists.

Queries which are currently cached on disk are split into two variants: `query_name`(cached) and `query_name_(of|for)_const_arg`(not cached), with `query_name_of_const_arg` taking a pair `(did, param_did): (LocalDefId, DefId)`.

For some queries a method `query_name_of_opt_const_arg` is added to `TyCtxt` which takes a `ty::WithOptConstParam` and either calls `query_name` or `query_name_of_const_arg` depending on the value of `const_param_did`.

r? @eddyb @varkor
2020-07-15 12:49:25 +00:00
Bastian Kauschke
2666aed498 unify Instance::resolve 2020-07-15 13:06:47 +02:00
Bastian Kauschke
e070b45e6a unsafety_check_result_for_const_arg 2020-07-15 13:06:47 +02:00
Bastian Kauschke
aca66bd052 WithOptConstParam::dummy -> WithOptConstParam::unknown 2020-07-15 13:06:47 +02:00
Bastian Kauschke
8003ccfdcd ty_def_id -> def_id_for_type_of 2020-07-15 13:06:47 +02:00
Bastian Kauschke
a909eb6b65 improve naming 2020-07-15 13:06:47 +02:00
Bastian Kauschke
805c44d5d3 cleanup 2020-07-15 12:58:33 +02:00
Bastian Kauschke
a7fe4df04a update promoted_mir 2020-07-15 12:58:33 +02:00
Bastian Kauschke
ae80d7e012 update const arg queries 2020-07-15 12:58:33 +02:00
Bastian Kauschke
08394eb121 update test 2020-07-15 12:58:32 +02:00
Bastian Kauschke
29b5844c2d only call typeck_tables_of_const_arg for const args 2020-07-15 12:58:32 +02:00
Bastian Kauschke
9df03ccf62 mir opt cross compile 2020-07-15 12:58:32 +02:00
Bastian Kauschke
d4cb820528 mir_built is a lie 2020-07-15 12:58:32 +02:00
Bastian Kauschke
aa02692570 decode stuff 2020-07-15 12:58:32 +02:00
Bastian Kauschke
050acc0213 ui test diff 2020-07-15 12:58:32 +02:00
Bastian Kauschke
4ef1029f5c mir opt diff 2020-07-15 12:58:32 +02:00
Bastian Kauschke
48bbf49765 const generics work! 2020-07-15 12:58:32 +02:00
Bastian Kauschke
b615b98f33 continue mir pipeline 2020-07-15 12:58:32 +02:00
Bastian Kauschke
316128c38a optimized_mir 2020-07-15 12:58:32 +02:00
Bastian Kauschke
a8419099d1 InstanceDef::Item 2020-07-15 12:58:32 +02:00
Bastian Kauschke
178c6507f6 typeck all the tables 2020-07-15 12:58:32 +02:00
Bastian Kauschke
2e6bf0923b const_eval_resolve 2020-07-15 12:58:32 +02:00
Bastian Kauschke
58031c7cb2 ConstKind::Unevaluated 2020-07-15 12:58:32 +02:00
Bastian Kauschke
08865d94e5 begin using WithOptParam 2020-07-15 12:58:32 +02:00
Bastian Kauschke
37a6103e1a introduce the query opt_const_param_of 2020-07-15 12:58:32 +02:00
Bastian Kauschke
2111d6f276 add const generic tests 2020-07-15 12:58:32 +02:00
bors
d9e8d62907 Auto merge of #74214 - nnethercote:change-SymbolName-name, r=eddyb
Change `SymbolName::name` to a `&str`.

This eliminates a bunch of `Symbol::intern()` and `Symbol::as_str()`
calls, which is good, because they require locking the interner.

Note that the unsafety in `from_cycle_error()` is identical to the
unsafety on other adjacent impls.

r? @eddyb
2020-07-15 08:44:46 +00:00
Bastian Kauschke
fcf52c167f improve DiscriminantKind handling
This now reuses `fn discriminant_ty` in project, removing
some code duplication. Doing so made me realize that
we previously had a disagreement about the discriminant
type of generators, with MIR using `u32` and codegen and
trait selection using `i32`.

We now always use `u32`.
2020-07-15 10:20:44 +02:00
bors
c714eae0e3 Auto merge of #71272 - jclulow:illumos-x86-ci, r=pietroalbini
build dist for x86_64-unknown-illumos

This change creates a new Docker image, "dist-x86_64-illumos", and sets
things up to build the full set of "dist" packages for illumos hosts, so
that illumos users can use "rustup" to install packages.  It also
adjusts the manifest builder to expect complete toolchains for this
platform.
2020-07-15 04:46:06 +00:00
Nicholas Nethercote
a1b8540390 Undo the const_str changes from the previous commit. 2020-07-15 14:38:00 +10:00
Nicholas Nethercote
7c4d431c06 Change SymbolName::name to a &str.
This eliminates a bunch of `Symbol::intern()` and `Symbol::as_str()`
calls, which is good, because they require locking the interner.

Note that the unsafety in `from_cycle_error()` is identical to the
unsafety on other adjacent impls.
2020-07-15 14:37:55 +10:00
Esteban Küber
ce3bd29c68 Handle case of incomplete local ty more gracefully
When encountering a local binding with a type that isn't completed, the
parser will reach a `=` token. When this happen, consider the type
"complete" as far as the parser is concerned to avoid further errors
being emitted by parse recovery logic.
2020-07-14 18:30:44 -07:00
bors
567ad7455d Auto merge of #74175 - nnethercote:more-static-symbols, r=oli-obk
More static symbols

These commits add some more static symbols and convert lots of places to use them.

r? @oli-obk
2020-07-15 00:16:25 +00:00
Esteban Küber
c44ca17a44 Remove string comparison and use diagnostic item instead 2020-07-14 16:08:02 -07:00
Nicholas Nethercote
5930081f34 Remove lots of Symbol::as_str() calls.
In various ways, such as changing functions to take a `Symbol` instead
of a `&str`.
2020-07-15 09:01:35 +10:00
Nicholas Nethercote
f04e866e57 Add and use more static symbols.
Note that the output of `unpretty-debug.stdout` has changed. In that
test the hash values are normalized from a symbol numbers to small
numbers like "0#0" and "0#1". The increase in the number of static
symbols must have caused the original numbers to contain more digits,
resulting in different pretty-printing prior to normalization.
2020-07-15 08:42:59 +10:00