Commit Graph

23359 Commits

Author SHA1 Message Date
Wesley Wiser
525d0dd6e2 Factor out debuginfo offset calculation 2022-12-08 20:38:23 -05:00
Ramon de C Valle
65698ae9f3 Add LLVM KCFI support to the Rust compiler
This commit adds LLVM Kernel Control Flow Integrity (KCFI) support to
the Rust compiler. It initially provides forward-edge control flow
protection for operating systems kernels for Rust-compiled code only by
aggregating function pointers in groups identified by their return and
parameter types. (See llvm/llvm-project@cff5bef.)

Forward-edge control flow protection for C or C++ and Rust -compiled
code "mixed binaries" (i.e., for when C or C++ and Rust -compiled code
share the same virtual address space) will be provided in later work as
part of this project by identifying C char and integer type uses at the
time types are encoded (see Type metadata in the design document in the
tracking issue #89653).

LLVM KCFI can be enabled with -Zsanitizer=kcfi.

Co-authored-by: bjorn3 <17426603+bjorn3@users.noreply.github.com>
2022-12-08 17:24:39 -08:00
pkubaj
32c777e891 Fix build on powerpc-unknown-freebsd
Probably also fixes build on mips*. Related to https://github.com/rust-lang/rust/issues/104220
2022-12-09 02:11:15 +01:00
Esteban Küber
132a140214 Point at LHS on binop type err if relevant 2022-12-08 15:30:57 -08:00
Maybe Waffle
c44c82de2b Assert size of rustc_parse_format::Piece<'_> 2022-12-08 22:50:39 +00:00
Arpad Borsos
ecf812777a Fix Async Generator ABI
This change was missed when making async generators implement `Future` directly.
It did not cause any problems in codegen so far, as `GeneratorState<(), Output>`
happens to have the same ABI as `Poll<Output>`.
2022-12-08 23:27:57 +01:00
Dan Gohman
98ae83daae Mangle "main" as "__main_void" on wasm32-wasi
On wasm, the age-old C trick of having a main function which can either have
no arguments or argc+argv doesn't work, because wasm requires caller and
callee signatures to match. WASI's current strategy is to have compilers
mangle main's name to indicate which signature they're using. Rust uses the
no-argument form, which should be mangled as `__main_void`.

This is needed on wasm32-wasi as of #105395.
2022-12-08 13:15:40 -08:00
Gary Guo
6275b045d7 Move paren expr and loop HIR lowering 2022-12-08 17:28:21 +00:00
Gary Guo
0c0fb226bb Support #[track_caller] on async closures 2022-12-08 17:23:44 +00:00
Santiago Pastorino
99d229095e Make encode_info_for_trait_item use queries instead of accessing the HIR 2022-12-08 11:58:14 -03:00
Matthias Krüger
660795eee5 Rollup merge of #105441 - nnethercote:rm-UnsafetyState, r=lcnr
Remove `UnsafetyState`

r? `@lcnr`
2022-12-08 12:57:33 +01:00
Matthias Krüger
fbfc5ada02 Rollup merge of #105423 - oli-obk:symbols, r=jackh726
Use `Symbol` for the crate name instead of `String`/`str`

It always got converted to a symbol anyway
2022-12-08 12:57:32 +01:00
Matthias Krüger
2fbde2b028 Rollup merge of #105408 - cuviper:help-rpitirpit, r=compiler-errors
Add help for `#![feature(impl_trait_in_fn_trait_return)]`

This adds a new variant `ImplTraitContext::FeatureGated`, so we can
generalize the help for `return_position_impl_trait_in_trait` to also
work for `impl_trait_in_fn_trait_return`.

cc #99697
2022-12-08 12:57:31 +01:00
Matthias Krüger
4d5a2f3d81 Rollup merge of #105405 - sunfishcode:sunfishcode/export-dynamic, r=TaKO8Ki
Stop passing -export-dynamic to wasm-ld.

-export-dynamic was a temporary hack added in the early days of the Rust wasm32 target when Rust didn't have a way to specify wasm exports in the source code. This flag causes all global symbols, and some compiler-internal symbols, to be exported, which is often more than needed.

Rust now does have a way to specify exports in the source code: `#[export_name = "..."]`.

So as the original comment suggests, -export-dynamic can now be removed, allowing users to have smaller binaries and better encapsulation in their wasm32-unknown-unknown modules.

It's possible that this change will require existing wasm32-unknown-unknown users will to add explicit `#[export_name = "..."]` directives to exporrt the symbols that their programs depend on having exported.
2022-12-08 12:57:30 +01:00
Matthias Krüger
f1f7560598 Rollup merge of #105317 - RalfJung:retag-rework, r=oli-obk
make retagging work even with 'unstable' places

This is based on top of https://github.com/rust-lang/rust/pull/105301. Only the last two commits are new.

While investigating https://github.com/rust-lang/unsafe-code-guidelines/issues/381 I realized that we would have caught this issue much earlier if the add_retag pass wouldn't bail out on assignments of the form `*ptr = ...`.

So this PR changes our retag strategy:
- When a new reference is created via `Rvalue::Ref` (or a raw ptr via `Rvalue::AddressOf`), we do the retagging as part of just executing that address-taking operation.
- For everything else, we still insert retags -- these retags basically serve to ensure that references stored in local variables (and their fields) are always freshly tagged, so skipping this for assignments like `*ptr = ...` is less egregious.
r? ```@oli-obk```
2022-12-08 12:57:30 +01:00
Matthias Krüger
e826a9acb4 Rollup merge of #105255 - cjgillot:issue-105197, r=compiler-errors
Make nested RPIT inherit the parent opaque's generics.

Fixes https://github.com/rust-lang/rust/issues/105197

r? ```@compiler-errors```
2022-12-08 12:57:29 +01:00
Matthias Krüger
086bdbbd73 Rollup merge of #104922 - estebank:fur-elize, r=oli-obk
Detect long types in E0308 and write them to disk

On type error with long types, print an abridged type and write the full type to disk.

Print the widest possible short type while still fitting in the terminal.
2022-12-08 12:57:28 +01:00
lcnr
dd9d05cec4 validate: use the correct reveal during opts 2022-12-08 11:24:25 +01:00
Erik Desjardins
a99e97af97 Add 0..=isize::MAX range metadata to size loads from vtables 2022-12-08 01:30:07 -05:00
Michael Goulet
3b9daac6a2 Move some suggestions from error_reporting to error_reporting::suggest 2022-12-08 05:58:30 +00:00
Michael Goulet
25a6daccab Move codegen_select_candidate to a rustc_traits 2022-12-08 05:16:57 +00:00
Michael Goulet
b36035c20f Move vtable methods into its own module 2022-12-08 05:16:57 +00:00
Michael Goulet
48270044b0 Move has_structural_eq_impls provider to rustc_ty_utils 2022-12-08 05:16:48 +00:00
Takayuki Maeda
85d7d9b6b7 add a test case for generic_const_exprs in trait items 2022-12-08 14:11:29 +09:00
Michael Goulet
7d23e29f9f Pull out logic into distinct functions 2022-12-08 04:59:10 +00:00
Nicholas Nethercote
9af48e5ab6 Fix a typo. 2022-12-08 15:53:31 +11:00
Nicholas Nethercote
7a4669285a Remove UnsafetyState.
`FnCtxt::ps` is the only occurrence. It gets updated during HIR
typechecking, but is never looked at.
2022-12-08 15:42:57 +11:00
Nicholas Nethercote
68a19209e0 Inline and remove place_contents_drop_state_cannot_differ.
It has a single call site and is hot enough to be worth inlining. And
make sure `is_terminal_path` is inlined, too.
2022-12-08 13:59:29 +11:00
Oli Scherer
d30848b30a Use Symbol for the crate name instead of String/str 2022-12-07 20:30:02 +00:00
Michael Goulet
57b722688d Properly print generator interior type sizes 2022-12-07 18:33:26 +00:00
Michael Goulet
b0dcadfc45 Move closure/generator type info methods to TyCtxt 2022-12-07 17:00:33 +00:00
Alex Brachet
5d88d36053 Don't internalize __llvm_profile_counter_bias
Currently, LLVM profiling runtime counter relocation cannot be
used by rust during LTO because symbols are being internalized
before all symbol information is known.

This mode makes LLVM emit a __llvm_profile_counter_bias symbol
which is referenced by the profiling initialization, which itself
is pulled in by the rust driver here [1].

It is enabled with -Cllvm-args=-runtime-counter-relocation for
platforms which are opt-in to this mode like Linux. On these
platforms there will be no link error, rather just surprising
behavior for a user which request runtime counter relocation.
The profiling runtime will not see that symbol go on as if it
were never there. On Fuchsia, the profiling runtime must have
this symbol which will cause a hard link error.

As an aside, I don't have enough context as to why rust's LTO
model is how it is. AFAICT, the internalize pass is only safe
to run at link time when all symbol information is actually
known, this being an example as to why. I think special casing
this symbol as a known one that LLVM can emit which should not
have it's visbility de-escalated should be fine given how
seldom this pattern of defining an undefined symbol to get
initilization code pulled in is. From a quick grep,
__llvm_profile_runtime is the only symbol that rustc does this
for.

[1] 0265a3e93b/compiler/rustc_codegen_ssa/src/back/linker.rs (L598)
2022-12-07 16:32:59 +00:00
Matthias Krüger
4968af0ee8 Rollup merge of #105418 - BelovDV:fix-libc-hack, r=petrochenkov
fix: remove hack from link.rs

This logic implemented in libc.

r? `@petrochenkov`
2022-12-07 15:39:09 +01:00
Matthias Krüger
b23419b03d Rollup merge of #105400 - BoxyUwU:braced_param_evaluatability, r=oli-obk
normalize before handling simple checks for evaluatability of `ty::Const`

`{{{{{{{ N }}}}}}}` is desugared into a `ConstKind::Unevaluated` for an anonymous `const` item so when calling `is_const_evaluatable` on it we skip the `ConstKind::Param(_) => Ok(())` arm which is incorrect.
2022-12-07 15:39:08 +01:00
Matthias Krüger
52cec8c99f Rollup merge of #105368 - WaffleLapkin:deref-even-harder, r=TaKO8Ki
Remove more `ref` patterns from the compiler

Previous PR: #105045
2022-12-07 15:39:07 +01:00
Matthias Krüger
95da525982 Rollup merge of #105343 - nbdd0121:hir, r=fee1-dead
Simplify attribute handling in rustc_ast_lowering

Given that attributes is stored in a separate BTreeMap, it's not necessary to pass it in when constructing `hir::Expr`. We can just construct `hir::Expr` and then call `self.lower_attrs` later if it needs attributes.

As most desugaring code don't use attributes, this allows some code cleanup.
2022-12-07 15:39:07 +01:00
Matthias Krüger
3bcfa4c459 Rollup merge of #105267 - compiler-errors:issue-104613, r=oli-obk
Don't ICE in ExprUseVisitor on FRU for non-existent struct

Fixes #104613
Fixes #105202
2022-12-07 15:39:06 +01:00
bors
91b8f34ac2 Auto merge of #104799 - pcc:linkage-fn, r=tmiasko
Support Option and similar enums as type of static variable with linkage attribute

Compiler MCP:
rust-lang/compiler-team#565
2022-12-07 10:24:59 +00:00
Daniil Belov
a9cf163c08 fix: remove hack from link.rs (moved to libc) 2022-12-07 13:13:58 +03:00
Nicholas Nethercote
d049be30cf Split EarlyContextAndPasses::check_id in two. 2022-12-07 19:29:31 +11:00
Nicholas Nethercote
b08fd6e8ef Inline some hot lint pass functions.
These each have a single call site, due to being called from a
"combined" lint pass.
2022-12-07 19:29:31 +11:00
Nicholas Nethercote
ec117c0ebd Rename $passes as $pass in several macros.
Because it makes more sense that way.
2022-12-07 19:29:31 +11:00
Nicholas Nethercote
ac1e69c903 Add some useful comments. 2022-12-07 19:29:31 +11:00
Nicholas Nethercote
0c844730d3 Remove $hir argument from late_lint_methods!.
Because it's always `'tcx`.
2022-12-07 19:29:26 +11:00
yukang
5599f2ad09 fix #105226, Detect spurious ; before assoc fn body 2022-12-07 16:13:39 +08:00
Takayuki Maeda
cb596e3015 consider parent_count for const param defaults 2022-12-07 12:55:30 +09:00
bors
e491d2b24d Auto merge of #105291 - nnethercote:remove-LintPassObjects, r=cjgillot
Remove `{Early,Late}LintPassObjects`.

`EarlyContextAndPass` wraps a single early lint pass. We aggregate multiple passes into that single pass by using `EarlyLintPassObjects`.

This commit removes `EarlyLintPassObjects` by changing `EarlyContextAndPass` into `EarlyContextAndPasses`. I.e. it just removes a level of indirection. This makes the code simpler and slightly faster.

The commit does likewise for late lints.

r? `@cjgillot`
2022-12-07 03:52:19 +00:00
Michael Goulet
b45b9489bb Compute generator sizes with -Zprint_type_sizes 2022-12-07 03:10:56 +00:00
Josh Stone
e9dd59131b Add help for #![feature(impl_trait_in_fn_trait_return)]
This adds a new variant `ImplTraitContext::FeatureGated`, so we can
generalize the help for `return_position_impl_trait_in_trait` to also
work for `impl_trait_in_fn_trait_return`.
2022-12-06 17:53:50 -08:00
Dan Gohman
3a07aa9b5e Stop passing -export-dynamic to wasm-ld.
-export-dynamic was a temporary hack added in the early days of the Rust
wasm32 target when Rust didn't have a way to specify wasm exports in the
source code. This flag causes all global symbols, and some compiler-internal
symbols, to be exported, which is often more than needed.

Rust now does have a way to specify exports in the source code:
`#[export_name = "..."]`.

So as the original comment suggests, -export-dynamic can now be removed,
allowing users to have smaller binaries and better encapsulation in
their wasm32-unknown-unknown modules.

It's possible that this change will require existing wasm32-unknown-unknown
users will to add explicit `#[export_name = "..."]` directives to
exporrt the symbols that their programs depend on having exported.
2022-12-06 16:50:29 -08:00