74 Commits

Author SHA1 Message Date
Nilstrieb
e3d397a5f6 Make ExprKind the first field in thir::Expr
This makes its `Debug` impl print it first which is useful, as it's the
most important part when looking at an expr.
2023-03-02 18:21:44 +00:00
bors
6290ae92b2 Auto merge of #108487 - cjgillot:no-typeck-mir, r=oli-obk
Avoid invoking typeck from borrowck

This PR attempts to reduce direct dependencies between typeck and MIR-related queries. The goal is to have all the information transit either through THIR or through dedicated queries that avoid depending on the whole `TypeckResults`.

In a first commit, we store the type information that MIR building requires into THIR. This avoids edges between mir_built and typeck.

In the second and third commit, we wrap informations around closures (upvars, kind origin and user-provided signature) to avoid borrowck depending on typeck information.

There should be a single remaining borrowck -> typeck edge in the good path, due to inline consts.
2023-02-27 21:48:10 +00:00
Camille GILLOT
380ca0c9e6 Move THIR printing to rustc_mir_build. 2023-02-26 14:31:15 +00:00
Camille GILLOT
d35dbbdc8e Store the body type in THIR. 2023-02-26 10:30:27 +00:00
b-naber
90e4c13409 output tree representation for thir-tree 2023-01-26 11:31:55 +01:00
mejrs
5d2b9a9ed0
Migrate deconstruct_pat.rs 2023-01-11 14:39:49 -08:00
Nilstrieb
8bfd6450c7 A few small cleanups for newtype_index
Remove the `..` from the body, only a few invocations used it and it's
inconsistent with rust syntax.

Use `;` instead of `,` between consts. As the Rust syntax gods inteded.
2022-12-18 21:47:28 +01:00
Nilstrieb
d679764fb6 Make #[debug_format] an attribute in newtype_index
This removes the `custom` format functionality as its only user was
trivially migrated to using a normal format.

If a new use case for a custom formatting impl pops up, you can add it
back.
2022-12-18 21:37:38 +01:00
Vadim Petrochenkov
7a5376d23c Unreserve braced enum variants in value namespace 2022-11-21 22:40:06 +03:00
Nilstrieb
7bfef19844 Use tidy-alphabetical in the compiler 2022-10-12 17:49:10 +05:30
Guillaume Gomez
adc24d1b5e Fix compiler docs 2022-10-10 18:28:29 +02:00
Pietro Albini
3975d55d98
remove cfg(bootstrap) 2022-09-26 10:14:45 +02:00
bors
c524c7dd25 Auto merge of #98588 - b-naber:valtrees-cleanup, r=lcnr
Use only ty::Unevaluated<'tcx, ()> in type system

r? `@lcnr`
2022-09-17 03:04:22 +00:00
b-naber
6af8fb7936 address review again 2022-09-14 17:30:25 +02:00
Camille GILLOT
af128b0144 Also compute implicit params in THIR. 2022-09-13 19:18:24 +02:00
Camille GILLOT
445841cda3 Compute explicit MIR params on THIR. 2022-09-13 19:18:23 +02:00
Michael Benfield
d7a750b504 Use niche-filling optimization even when multiple variants have data.
Fixes #46213
2022-09-07 20:12:45 +00:00
Michael Goulet
be53bd8380 Include enum path in variant suggestion 2022-09-03 08:11:17 +00:00
Nicholas Nethercote
43a0268f3a Use boxed slices in PatKind.
To shrink it a little more.
2022-09-02 07:26:22 +10:00
Nicholas Nethercote
a40124e01c Box CanonicalUserTypeAnnotation::CanonicalUserType.
This shrinks `Ascription`, which shrinks `PatKind::AscribeUserType`,
which shrinks `Pat`.
2022-09-02 07:26:22 +10:00
Nicholas Nethercote
2c4c8eb1a3 Box PatKind::Range.
Because it's the biggest variant. Also, make `PatRange` non-`Copy`,
because it's 104 bytes, which is pretty big.
2022-09-02 07:26:22 +10:00
Nicholas Nethercote
053874eecc Clean up THIR patterns.
`thir::Pat::kind` is a `Box<PatKind>`, which doesn't follow the usual
pattern in AST/HIR/THIR which is that the "kind" enum for a node is
stored inline within the parent struct.

This commit makes the `PatKind` directly inline within the `Pat`. This
requires using `Box<Pat>` in all the types that hold a `Pat.

Ideally, `Pat` would be stored in `Thir` like `Expr` and `Stmt` and
referred to with a `PatId` rather than `Box<Pat>`. But this is hard to
do because lots of `Pat`s get created after the destruction of the `Cx`
that does normal THIR building. But this does get us a step closer to
`PatId`, because all the `Box<Pat>` occurrences would be replaced with
`PatId` if `PatId` ever happened.

At 128 bytes, `Pat` is large. Subsequent commits will shrink it.
2022-09-02 07:26:17 +10:00
Nicholas Nethercote
5a41eb81ca Add some more THIR size assertions. 2022-08-24 15:06:50 +10:00
Nicholas Nethercote
e3755c1d54 Rename thir::Adt as thir::AdtExpr.
This matches the naming scheme used elsewhere, e.g. in the AST, and
avoids name clashes with the `ExprKind::Closure` variant.
2022-08-24 15:06:50 +10:00
Nicholas Nethercote
e57ac764b8 Box thir::ExprKind::InlineAsm.
This shrinks `thir::Expr`.
2022-08-24 15:06:50 +10:00
Nicholas Nethercote
b3245a8dff Box thir::ExprKind::Closure.
This shrinks `thir::Expr`.
2022-08-24 15:06:50 +10:00
Nicholas Nethercote
2df805fc7a Store blocks in Thir.
Like expressions, statements, and match arms. This shrinks `thir::Stmt`
and is a precursor to further shrinking `thir::Expr`.
2022-08-24 15:06:44 +10:00
Nicholas Nethercote
e7c25c3a97 Box user_ty fields in thir::ExprKind.
This shrinks several large variants of `ExprKind`.
2022-08-24 14:16:09 +10:00
Nicholas Nethercote
c9429b1cec Define index types within thir_with_elements.
The macro already generates other stuff, might as well generate these
index types as well.
2022-08-24 11:37:56 +10:00
Nicholas Nethercote
574ba831d4 Avoid repeating qualifiers on static_assert_size calls.
Some of these don't need a `use` statement because there is already a
`#[macro_use] extern crate rustc_data_structures` item in the crate.
2022-08-10 11:51:21 +10:00
Matthias Krüger
e6bb00fff5
Rollup merge of #100003 - nnethercote:improve-size-assertions, r=lqd
Improve size assertions.

- For any file with four or more size assertions, move them into a
  separate module (as is already done for `hir.rs`).
- Add some more for AST nodes and THIR nodes.
- Put the `hir.rs` ones in alphabetical order.

r? `@lqd`
2022-08-01 16:49:33 +02:00
Nicholas Nethercote
9037ebba0c Improve size assertions.
- For any file with four or more size assertions, move them into a
  separate module (as is already done for `hir.rs`).
- Add some more for AST nodes and THIR nodes.
- Put the `hir.rs` ones in alphabetical order.
2022-08-01 09:15:05 +10:00
Cameron Steffen
cf2433a74f Use LocalDefId for closures more 2022-07-30 15:59:17 -05:00
Dylan DPC
ecae3d74e2
Rollup merge of #99000 - JulianKnodt:allow_resolve_no_substs, r=lcnr
Move abstract const to middle

Moves AbstractConst (and all associated methods) to rustc middle for use in `rustc_infer`.
This allows for const resolution in infer to use abstract consts to walk consts and check if
they are resolvable.

This attempts to resolve the issue where `Foo<{ concrete const }, generic T>` is incorrectly marked as conflicting, and is independent from the other issue where nested abstract consts must be resolved.

r? `@lcnr`
2022-07-14 19:24:04 +05:30
kadmin
e612e2603c Move abstract const to rustc_middle::ty 2022-07-12 02:21:31 +00:00
Ding Xiang Fei
6c529ded86
lower let-else in MIR instead 2022-07-11 23:20:36 +02:00
Ralf Jung
ac265cdc19 review feedback 2022-07-09 07:27:29 -04:00
Ralf Jung
a422b42159 don't allow ZST in ScalarInt
There are several indications that we should not ZST as a ScalarInt:
- We had two ways to have ZST valtrees, either an empty `Branch` or a `Leaf` with a ZST in it.
  `ValTree::zst()` used the former, but the latter could possibly arise as well.
- Likewise, the interpreter had `Immediate::Uninit` and `Immediate::Scalar(Scalar::ZST)`.
- LLVM codegen already had to special-case ZST ScalarInt.

So instead add new ZST variants to those types that did not have other variants
which could be used for this purpose.
2022-07-09 07:27:29 -04:00
Alan Egerton
e4b9625b87
Add #[derive(TypeVisitable)] 2022-07-05 22:25:15 +01:00
Ding Xiang Fei
6cad569a95
thir: wrap hir id of vars into local var id 2022-06-09 17:00:38 +08:00
Tomasz Miąsko
0e7eca77e1 Fix precise field capture of univariant enums
When constructing a MIR from a THIR field expression, introduce an
additional downcast projection before accessing a field of an enum.

When rebasing a place builder on top of a captured place, account for
the fact that a single HIR enum field projection corresponds to two MIR
projection elements: a downcast element and a field element.
2022-05-23 19:07:06 +02:00
lcnr
39a03779f8 correctly deal with user type ascriptions in pat 2022-05-21 08:13:17 +02:00
b-naber
96b36d6eb2 use GlobalId in eval_to_valtree query and introduce query for valtree_to_const_val 2022-05-16 15:58:15 +02:00
Xavier Denis
eed91ee8b3 Make all thir types implement clone 2022-04-20 00:11:08 +02:00
Amanieu d'Antras
dc345d8bff Reimplement lowering of sym operands for asm! so that it also works with global_asm! 2022-04-14 15:32:03 +01:00
b-naber
0078e54185 rebase and use ty::Const in patterns again 2022-04-02 12:21:00 +02:00
b-naber
ac60db231c do use ty::Const in patterns and abstract consts 2022-04-02 12:20:59 +02:00
b-naber
b38077ea0b change thir to use mir::ConstantKind instead of ty::Const 2022-04-02 12:20:56 +02:00
Yuri Astrakhan
5160f8f843 Spellchecking compiler comments
This PR cleans up the rest of the spelling mistakes in the compiler comments. This PR does not change any literal or code spelling issues.
2022-03-30 15:14:15 -04:00
b-naber
5fcccd1739 use NonHirLiteral instead of ScalarLiteral, move pattern related code to pat_is_poly in IsThirPolymorphic 2022-03-23 11:34:33 +01:00