Commit Graph

2404 Commits

Author SHA1 Message Date
bjorn3
fc0c753c2d Remove src_files and remove_file
They only apply to the main source archive and their role can be
fulfilled through the skip argument of add_archive too.
2022-06-14 15:11:14 +00:00
b-naber
9096b3e44f implement valtrees as the type-system representation for constant values 2022-06-14 16:07:11 +02:00
Nicholas Nethercote
ce2b3a9b4c Rename the ConstS::val field as kind.
And likewise for the `Const::val` method.

Because its type is called `ConstKind`. Also `val` is a confusing name
because `ConstKind` is an enum with seven variants, one of which is
called `Value`. Also, this gives consistency with `TyS` and `PredicateS`
which have `kind` fields.

The commit also renames a few `Const` variables from `val` to `c`, to
avoid confusion with the `ConstKind::Value` variant.
2022-06-14 13:06:44 +10:00
bors
b867d41ed4 Auto merge of #97825 - Dylan-DPC:rollup-ya51k1k, r=Dylan-DPC
Rollup of 5 pull requests

Successful merges:

 - #97058 (Various refactors to the incr comp workproduct handling)
 - #97301 (Allow unstable items to be re-exported unstably without requiring the feature be enabled)
 - #97738 (Fix ICEs from zsts within unsized types with non-zero offsets)
 - #97771 (Remove SIGIO reference on Haiku)
 - #97808 (Add some unstable target features for the wasm target codegen)

Failed merges:

r? `@ghost`
`@rustbot` modify labels: rollup
2022-06-07 11:08:58 +00:00
bors
d8bd0a950d Auto merge of #97512 - scottmcm:add-coldcc, r=nagisa,lcnr
Add support for emitting functions with `coldcc` to LLVM

The eventual goal is to try using this for things like the internal panicking stuff, to see whether it helps.
2022-06-07 08:12:45 +00:00
bjorn3
3d8e854604 Make saved_file field of WorkProduct non-optional
A WorkProduct without a saved file is useless
2022-06-06 12:39:32 +00:00
bjorn3
bbb850931f Factor Option out of copy_cgu_workproduct_to_incr_comp_cache_dir call
This improves clarity of the code a bit
2022-06-06 12:38:38 +00:00
bjorn3
5a1a111b52 Avoid an unnecessary clone for copy_cgu_workproduct_to_incr_comp_cache_dir calls 2022-06-06 12:32:08 +00:00
bjorn3
d5cb2bee82 Rename CodegenUnit::work_product to previous_work_product
It returns the previous work product or panics if there is none. This rename
makes the purpose of this method clearer.
2022-06-06 12:30:10 +00:00
Nikolai Vazquez
eb5f23737b Fix unsized field order 2022-06-03 12:58:36 -04:00
Nikolai Vazquez
7a8b96f89d Make std::mem::needs_drop accept ?Sized 2022-06-03 03:28:19 -04:00
Ralf Jung
ede985e8b8 add cast kind of from_exposed_addr (int-to-ptr casts) 2022-06-02 10:46:13 -04:00
Ralf Jung
722c724a95 rename PointerAddress → PointerExposeAddress 2022-06-01 14:08:17 -04:00
Tomasz Miąsko
e60087505e Add a pointer to address cast kind
A pointer to address cast are often special-cased.
Introduce a dedicated cast kind to make them easy distinguishable.
2022-05-31 00:00:00 +00:00
Scott McMurray
4ee48c09ee Use CallConv::Cold in cranelift for extern "rust-cold" 2022-05-30 00:19:24 -07:00
Scott McMurray
e2f90f7f74 Add support for emitting functions with coldcc in LLVM
The eventual goal is to try using this for things like the internal panicking stuff, to see whether it helps.
2022-05-30 00:19:23 -07:00
5225225
d27ec6c86d Add flag for stricter checks on uninit/zeroed 2022-05-24 14:26:52 +01:00
Jakob Degen
5011ae46a1 Refactor call terminator to always hold a destination place 2022-05-23 17:49:04 -04:00
bjorn3
d9025cebac Merge commit '63734fcdd718cca089f84c42f3a42c0096cfd431' into sync_cg_clif-2022-05-15 2022-05-15 12:32:19 +02:00
Scott McMurray
bb7b844d7c Rename unsigned_offset_from to sub_ptr 2022-05-11 17:16:25 -07:00
Scott McMurray
06817d1bb3 Add unsigned_offset_from on pointers
Like we have `add`/`sub` which are the `usize` version of `offset`, this adds the `usize` equivalent of `offset_from`.  Like how `.add(d)` replaced a whole bunch of `.offset(d as isize)`, you can see from the changes here that it's fairly common that code actually knows the order between the pointers and *wants* a `usize`, not an `isize`.

As a bonus, this can do `sub nuw`+`udiv exact`, rather than `sub`+`sdiv exact`, which can be optimized slightly better because it doesn't have to worry about negatives.  That's why the slice iterators weren't using `offset_from`, though I haven't updated that code in this PR because slices are so perf-critical that I'll do it as its own change.

This is an intrinsic, like `offset_from`, so that it can eventually be allowed in CTFE.  It also allows checking the extra safety condition -- see the test confirming that CTFE catches it if you pass the pointers in the wrong order.
2022-05-11 17:16:25 -07:00
bjorn3
32202f20cd Merge commit 'f2cdd4a78d89c009342197cf5844a21f8aa813df' into sync_cg_clif-2022-04-22 2022-04-22 21:11:38 +02:00
Jakob Degen
a923e9211a Add new Deinit statement kind 2022-04-11 08:55:03 -04:00
klensy
f2d2ec3e93 check_doc_keyword: don't alloc string for emptiness check
check_doc_alias_value: get argument as Symbol to prevent needless string convertions

check_doc_attrs: don't alloc vec, iterate over slice. Vec introduced in #83149, but no perf run posted on merge

replace as_str() check with symbol check

get_single_str_from_tts: don't prealloc string

trivial string to str replace

LifetimeScopeForPath::NonElided use Vec<Symbol> instead of Vec<String>

AssertModuleSource use BTreeSet<Symbol> instead of BTreeSet<String>

CrateInfo.crate_name replace FxHashMap<CrateNum, String> with FxHashMap<CrateNum, Symbol>
2022-04-08 11:45:57 +03:00
Oli Scherer
2efede7749 Mark scalar layout unions so that backends that do not support partially initialized scalars can special case them. 2022-04-05 13:18:21 +00:00
Loïc BRANSTETT
69046fa8fd Cleanup after some refactoring in rustc_target 2022-04-03 21:29:57 +02:00
Yuri Astrakhan
7665d87638 Addressed comments by @compiler-errors and @bjorn3 2022-03-30 17:04:46 -04:00
Yuri Astrakhan
421baaff3f 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
Dylan DPC
804cae3f9b Rollup merge of #95461 - nyurik:spelling, r=lcnr
Spellchecking some comments

This PR attempts to clean up some minor spelling mistakes in comments
2022-03-30 09:10:07 +02:00
Yuri Astrakhan
ae2f203e24 Spellchecking compiler code
Address some spelling mistakes in strings, private function names, and function params.
2022-03-30 01:42:10 -04:00
Yuri Astrakhan
93aedb67df Spellchecking some comments
This PR attempts to clean up some minor spelling mistakes in comments
2022-03-30 01:39:38 -04:00
bjorn3
1bd90f8644 Don't declare test_variadic_fnptr with two conflicting signatures
It is UB for LLVM and results in a compile error for Cranelift
2022-03-20 21:09:35 +01:00
bjorn3
d1904e5855 Merge commit '370c397ec9169809e5ad270079712e0043514240' into sync_cg_clif-2022-03-20 2022-03-20 16:55:21 +01:00
bors
86daae2e6b Auto merge of #88098 - Amanieu:oom_panic, r=nagisa
Implement -Z oom=panic

This PR removes the `#[rustc_allocator_nounwind]` attribute on `alloc_error_handler` which allows it to unwind with a panic instead of always aborting. This is then used to implement `-Z oom=panic` as per RFC 2116 (tracking issue #43596).

Perf and binary size tests show negligible impact.
2022-03-18 03:01:46 +00:00
mark
c56a10ca3b rustc_error: make ErrorReported impossible to construct
There are a few places were we have to construct it, though, and a few
places that are more invasive to change. To do this, we create a
constructor with a long obvious name.
2022-03-16 10:35:24 -05:00
Nicholas Nethercote
e439761b8f Improve AdtDef interning.
This commit makes `AdtDef` use `Interned`. Much the commit is tedious
changes to introduce getter functions. The interesting changes are in
`compiler/rustc_middle/src/ty/adt.rs`.
2022-03-11 13:31:24 +11:00
Nicholas Nethercote
8723fe0b6b Clarify Layout interning.
`Layout` is another type that is sometimes interned, sometimes not, and
we always use references to refer to it so we can't take any advantage
of the uniqueness properties for hashing or equality checks.

This commit renames `Layout` as `LayoutS`, and then introduces a new
`Layout` that is a newtype around an `Interned<LayoutS>`. It also
interns more layouts than before. Previously layouts within layouts
(via the `variants` field) were never interned, but now they are. Hence
the lifetime on the new `Layout` type.

Unlike other interned types, these ones are in `rustc_target` instead of
`rustc_middle`. This reflects the existing structure of the code, which
does layout-specific stuff in `rustc_target` while `TyAndLayout` is
generic over the `Ty`, allowing the type-specific stuff to occur in
`rustc_middle`.

The commit also adds a `HashStable` impl for `Interned`, which was
needed. It hashes the contents, unlike the `Hash` impl which hashes the
pointer.
2022-03-07 13:41:47 +11:00
Nicholas Nethercote
7c1a318c7b Introduce ConstAllocation.
Currently some `Allocation`s are interned, some are not, and it's very
hard to tell at a use point which is which.

This commit introduces `ConstAllocation` for the known-interned ones,
which makes the division much clearer. `ConstAllocation::inner()` is
used to get the underlying `Allocation`.

In some places it's natural to use an `Allocation`, in some it's natural
to use a `ConstAllocation`, and in some places there's no clear choice.
I've tried to make things look as nice as possible, while generally
favouring `ConstAllocation`, which is the type that embodies more
information. This does require quite a few calls to `inner()`.

The commit also tweaks how `PartialOrd` works for `Interned`. The
previous code was too clever by half, building on `T: Ord` to make the
code shorter. That caused problems with deriving `PartialOrd` and `Ord`
for `ConstAllocation`, so I changed it to build on `T: PartialOrd`,
which is slightly more verbose but much more standard and avoided the
problems.
2022-03-07 08:25:50 +11:00
Amanieu d'Antras
e29149552e Add -Z oom={panic,abort} command-line option 2022-03-03 12:58:38 +00:00
mark
346108202d rename ErrorReported -> ErrorGuaranteed 2022-03-02 09:45:25 -06:00
Tomasz Miąsko
69bd5324ca Normalize main return type during mono item collection & codegen 2022-02-23 22:33:50 +01:00
bjorn3
b5cbb87e62 Merge commit '35d9c6bf256968e1b40e0d554607928bdf9cebea' into sync_cg_clif-2022-02-23 2022-02-23 11:49:34 +01:00
Matthias Krüger
a063e138b6 Rollup merge of #94169 - Amanieu:asm_stuff, r=nagisa
Fix several asm! related issues

This is a combination of several fixes, each split into a separate commit. Splitting these into PRs is not practical since they conflict with each other.

Fixes #92378
Fixes #85247

r? ``@nagisa``
2022-02-22 12:16:28 +01:00
Amanieu d'Antras
73cf3aaa78 Take CodegenFnAttrs into account when validating asm! register operands
Checking of asm! register operands now properly takes function
attributes such as #[target_feature] and #[instruction_set] into
account.
2022-02-21 18:28:22 +00:00
Amanieu d'Antras
e62c26e39d On ARM, use relocation_model to detect whether r9 should be reserved
The previous approach of checking for the reserve-r9 target feature
didn't actually work because LLVM only sets this feature very late when
initializing the per-function subtarget.
2022-02-21 18:28:22 +00:00
lcnr
d34bcdd49c use List<Ty<'tcx>> for tuples 2022-02-21 07:09:11 +01:00
Mark Rousskov
7e80bc3c8d Move ty::print methods to Drop-based scope guards 2022-02-16 17:24:23 -05:00
Nicholas Nethercote
06bc64df92 Overhaul Const.
Specifically, rename the `Const` struct as `ConstS` and re-introduce `Const` as
this:
```
pub struct Const<'tcx>(&'tcx Interned<ConstS>);
```
This now matches `Ty` and `Predicate` more closely, including using
pointer-based `eq` and `hash`.

Notable changes:
- `mk_const` now takes a `ConstS`.
- `Const` was copy, despite being 48 bytes. Now `ConstS` is not, so need a
  we need separate arena for it, because we can't use the `Dropless` one any
  more.
- Many `&'tcx Const<'tcx>`/`&Const<'tcx>` to `Const<'tcx>` changes
- Many `ct.ty` to `ct.ty()` and `ct.val` to `ct.val()` changes.
- Lots of tedious sigil fiddling.
2022-02-15 16:19:59 +11:00
Nicholas Nethercote
18e7b7ece1 Overhaul RegionKind and Region.
Specifically, change `Region` from this:
```
pub type Region<'tcx> = &'tcx RegionKind;
```
to this:
```
pub struct Region<'tcx>(&'tcx Interned<RegionKind>);
```

This now matches `Ty` and `Predicate` more closely.

Things to note
- Regions have always been interned, but we haven't been using pointer-based
  `Eq` and `Hash`. This is now happening.
- I chose to impl `Deref` for `Region` because it makes pattern matching a lot
  nicer, and `Region` can be viewed as just a smart wrapper for `RegionKind`.
- Various methods are moved from `RegionKind` to `Region`.
- There is a lot of tedious sigil changes.
- A couple of types like `HighlightBuilder`, `RegionHighlightMode` now have a
  `'tcx` lifetime because they hold a `Ty<'tcx>`, so they can call `mk_region`.
- A couple of test outputs change slightly, I'm not sure why, but the new
  outputs are a little better.
2022-02-15 16:08:52 +11:00
Nicholas Nethercote
6a20fa93b5 Overhaul TyS and Ty.
Specifically, change `Ty` from this:
```
pub type Ty<'tcx> = &'tcx TyS<'tcx>;
```
to this
```
pub struct Ty<'tcx>(Interned<'tcx, TyS<'tcx>>);
```
There are two benefits to this.
- It's now a first class type, so we can define methods on it. This
  means we can move a lot of methods away from `TyS`, leaving `TyS` as a
  barely-used type, which is appropriate given that it's not meant to
  be used directly.
- The uniqueness requirement is now explicit, via the `Interned` type.
  E.g. the pointer-based `Eq` and `Hash` comes from `Interned`, rather
  than via `TyS`, which wasn't obvious at all.

Much of this commit is boring churn. The interesting changes are in
these files:
- compiler/rustc_middle/src/arena.rs
- compiler/rustc_middle/src/mir/visit.rs
- compiler/rustc_middle/src/ty/context.rs
- compiler/rustc_middle/src/ty/mod.rs

Specifically:
- Most mentions of `TyS` are removed. It's very much a dumb struct now;
  `Ty` has all the smarts.
- `TyS` now has `crate` visibility instead of `pub`.
- `TyS::make_for_test` is removed in favour of the static `BOOL_TY`,
  which just works better with the new structure.
- The `Eq`/`Ord`/`Hash` impls are removed from `TyS`. `Interned`s impls
  of `Eq`/`Hash` now suffice. `Ord` is now partly on `Interned`
  (pointer-based, for the `Equal` case) and partly on `TyS`
  (contents-based, for the other cases).
- There are many tedious sigil adjustments, i.e. adding or removing `*`
  or `&`. They seem to be unavoidable.
2022-02-15 16:03:24 +11:00