Commit graph

1024 commits

Author SHA1 Message Date
Ralf Jung
e8379c9598 interpret: get rid of 'mir lifetime everywhere 2024-05-27 08:25:57 +02:00
Ralf Jung
36d36a3e1f interpret: the MIR is actually at lifetime 'tcx 2024-05-27 07:45:41 +02:00
bors
8679004993 Auto merge of #125434 - nnethercote:rm-more-extern-tracing, r=jackh726
Remove more `#[macro_use] extern crate tracing`

Because explicit importing of macros via use items is nicer (more standard and readable) than implicit importing via `#[macro_use]`. Continuing the work from #124511 and #124914.

r? `@jackh726`
2024-05-23 21:36:54 +00:00
Oli Scherer
4cf34cb752 Allow const eval failures if the cause is a type layout issue 2024-05-23 10:51:52 +00:00
Nicholas Nethercote
7a5d814a04 Remove #[macro_use] extern crate tracing from rustc_const_eval. 2024-05-23 18:02:38 +10:00
bors
5293c6adb7 Auto merge of #125359 - RalfJung:interpret-overflowing-ops, r=oli-obk
interpret: make overflowing binops just normal binops

Follow-up to https://github.com/rust-lang/rust/pull/125173 (Cc `@scottmcm)`
2024-05-23 04:03:14 +00:00
bors
5d328a1f62 Auto merge of #117329 - RalfJung:offset-by-zero, r=oli-obk,scottmcm
offset: allow zero-byte offset on arbitrary pointers

As per prior `@rust-lang/opsem` [discussion](https://github.com/rust-lang/opsem-team/issues/10) and [FCP](https://github.com/rust-lang/unsafe-code-guidelines/issues/472#issuecomment-1793409130):

- Zero-sized reads and writes are allowed on all sufficiently aligned pointers, including the null pointer
- Inbounds-offset-by-zero is allowed on all pointers, including the null pointer
- `offset_from` on two pointers derived from the same allocation is always allowed when they have the same address

This removes surprising UB (in particular, even C++ allows "nullptr + 0", which we currently disallow), and it brings us one step closer to an important theoretical property for our semantics ("provenance monotonicity": if operations are valid on bytes without provenance, then adding provenance can't make them invalid).

The minimum LLVM we require (v17) includes https://reviews.llvm.org/D154051, so we can finally implement this.

The `offset_from` change is needed to maintain the equivalence with `offset`: if `let ptr2 = ptr1.offset(N)` is well-defined, then `ptr2.offset_from(ptr1)` should be well-defined and return N. Now consider the case where N is 0 and `ptr1` dangles: we want to still allow offset_from here.

I think we should change offset_from further, but that's a separate discussion.

Fixes https://github.com/rust-lang/rust/issues/65108
[Tracking issue](https://github.com/rust-lang/rust/issues/117945) | [T-lang summary](https://github.com/rust-lang/rust/pull/117329#issuecomment-1951981106)

Cc `@nikic`
2024-05-22 13:04:14 +00:00
Ralf Jung
cb5319483e clarify comment
Co-authored-by: scottmcm <scottmcm@users.noreply.github.com>
2024-05-22 11:19:04 +02:00
Ralf Jung
9526ce60fd improve comment wording 2024-05-21 21:13:20 +02:00
Ralf Jung
c0b4b454c3 interpret: make overflowing binops just normal binops 2024-05-21 14:50:09 +02:00
Scott McMurray
95c0e5c6a8 Remove Rvalue::CheckedBinaryOp 2024-05-17 20:33:02 -07:00
Ralf Jung
5c33a5690d offset, offset_from: allow zero-byte offset on arbitrary pointers 2024-05-13 07:59:16 +02:00
Nicholas Nethercote
4497d345a8 Remove extern crate rustc_middle from rustc_const_eval.
This requires exporting the interpreter macros so they can be used with
`use crate::interpret::*`.
2024-05-13 08:02:14 +10:00
Matthias Krüger
9a9ec90567
Rollup merge of #124957 - compiler-errors:builtin-deref, r=michaelwoerister
Make `Ty::builtin_deref` just return a `Ty`

Nowhere in the compiler are we using the mutability part of the `TyAndMut` that we used to return.
2024-05-10 16:10:47 +02:00
Michael Goulet
d50c2b0a52 Make builtin_deref just return a Ty 2024-05-09 22:55:00 -04:00
Ralf Jung
41d36a0951 interpret/miri: better errors on failing offset_from 2024-05-09 13:09:47 +02:00
Matthias Krüger
5d413c111a
Rollup merge of #124720 - RalfJung:interpret-drop, r=compiler-errors
interpret: Drop: always evaluate place

That way we can also avoid dealing with `instantiate_from_frame_and_normalize_erasing_regions`.
2024-05-04 22:27:33 +02:00
Ralf Jung
f0dee6bbe5 some comments or dynamic drop handling 2024-05-04 20:04:01 +02:00
Ralf Jung
86a933a574 interpret: Drop: always evaluate place 2024-05-04 19:59:11 +02:00
Ralf Jung
8e4466497f interpret, miri: uniform treatments of intrinsics/functions with and without return block 2024-05-04 17:39:29 +02:00
Matthias Krüger
ceb7b5e70e
Rollup merge of #124293 - oli-obk:miri_intrinsic_fallback_body, r=RalfJung
Let miri and const eval execute intrinsics' fallback bodies

fixes https://github.com/rust-lang/miri/issues/3397

r? ``@RalfJung``
2024-05-04 12:37:22 +02:00
Oli Scherer
351658ae66 Let miri and const eval execute intrinsics' fallback bodies 2024-05-03 09:01:12 +00:00
Ralf Jung
dba1849c22 interpret: hide some reexports in rustdoc 2024-05-02 18:47:36 +02:00
Ralf Jung
173d1bd36b properly fill a promoted's required_consts
then we can also make all_required_consts_are_checked a constant instead of a function
2024-04-23 23:02:54 +02:00
Ralf Jung
bf021ea625 interpret: sanity-check that required_consts captures all consts that can fail 2024-04-23 22:52:44 +02:00
Matthias Krüger
918304b190
Rollup merge of #124003 - WaffleLapkin:dellvmization, r=scottmcm,RalfJung,antoyo
Dellvmize some intrinsics (use `u32` instead of `Self` in some integer intrinsics)

This implements https://github.com/rust-lang/compiler-team/issues/693 minus what was implemented in #123226.

Note: I decided to _not_ change `shl`/... builder methods, as it just doesn't seem worth it.

r? ``@scottmcm``
2024-04-23 20:17:51 +02:00
Matthias Krüger
8039488e59
Rollup merge of #124220 - RalfJung:interpret-wrong-vtable, r=oli-obk
Miri: detect wrong vtables in wide pointers

Fixes https://github.com/rust-lang/miri/issues/3497.
Needed to catch the UB that https://github.com/rust-lang/rust/pull/123572 will start exploiting.

r? `@oli-obk`
2024-04-23 06:24:57 +02:00
bors
aca749eefc Auto merge of #121801 - zetanumbers:async_drop_glue, r=oli-obk
Add simple async drop glue generation

This is a prototype of the async drop glue generation for some simple types. Async drop glue is intended to behave very similar to the regular drop glue except for being asynchronous. Currently it does not execute synchronous drops but only calls user implementations of `AsyncDrop::async_drop` associative function and awaits the returned future. It is not complete as it only recurses into arrays, slices, tuples, and structs and does not have same sensible restrictions as the old `Drop` trait implementation like having the same bounds as the type definition, while code assumes their existence (requires a future work).

This current design uses a workaround as it does not create any custom async destructor state machine types for ADTs, but instead uses types defined in the std library called future combinators (deferred_async_drop, chain, ready_unit).

Also I recommend reading my [explainer](https://zetanumbers.github.io/book/async-drop-design.html).

This is a part of the [MCP: Low level components for async drop](https://github.com/rust-lang/compiler-team/issues/727) work.

Feature completeness:

 - [x] `AsyncDrop` trait
 - [ ] `async_drop_in_place_raw`/async drop glue generation support for
   - [x] Trivially destructible types (integers, bools, floats, string slices, pointers, references, etc.)
   - [x] Arrays and slices (array pointer is unsized into slice pointer)
   - [x] ADTs (enums, structs, unions)
   - [x] tuple-like types (tuples, closures)
   - [ ] Dynamic types (`dyn Trait`, see explainer's [proposed design](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#async-drop-glue-for-dyn-trait))
   - [ ] coroutines (https://github.com/rust-lang/rust/pull/123948)
 - [x] Async drop glue includes sync drop glue code
 - [x] Cleanup branch generation for `async_drop_in_place_raw`
 - [ ] Union rejects non-trivially async destructible fields
 - [ ] `AsyncDrop` implementation requires same bounds as type definition
 - [ ] Skip trivially destructible fields (optimization)
 - [ ] New [`TyKind::AdtAsyncDestructor`](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#adt-async-destructor-types) and get rid of combinators
 - [ ] [Synchronously undroppable types](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#exclusively-async-drop)
 - [ ] Automatic async drop at the end of the scope in async context
2024-04-23 02:10:23 +00:00
Scott McMurray
bb8d6f790b Address PR feedback 2024-04-21 11:08:37 -07:00
Scott McMurray
de64ff76f8 Use it in the library, and InstSimplify it away in the easy places 2024-04-21 11:08:37 -07:00
Ralf Jung
875f0c2da0 Miri: detect wrong vtables in wide pointers 2024-04-21 13:04:51 +02:00
bors
ce3263e60e Auto merge of #124113 - RalfJung:interpret-scalar-ops, r=oli-obk
interpret: use ScalarInt for bin-ops; avoid PartialOrd for ScalarInt

Best reviewed commit-by-commit

r? `@oli-obk`
2024-04-19 17:00:28 +00:00
Ralf Jung
42220f0930 ScalarInt: add methods to assert being a (u)int of given size 2024-04-19 13:51:52 +02:00
Ralf Jung
5e6184cdb7 interpret/binary_int_op: avoid dropping to raw ints until we determined the sign 2024-04-18 14:25:06 +02:00
bors
c25473ff62 Auto merge of #124008 - nnethercote:simpler-static_assert_size, r=Nilstrieb
Simplify `static_assert_size`s.

We want to run them on all 64-bit platforms.

r? `@ghost`
2024-04-18 09:47:45 +00:00
Nicholas Nethercote
0d97669a17 Simplify static_assert_sizes.
We want to run them on all 64-bit platforms.
2024-04-18 15:36:25 +10:00
bors
5260893724 Auto merge of #122684 - oli-obk:delay_interning_errors_to_after_validaiton, r=RalfJung
Delay interning errors to after validation

fixes https://github.com/rust-lang/rust/issues/122398
fixes #122548

This improves diagnostics since validation errors are usually more helpful compared with interning errors that just make broad statements about the entire constant

r? `@RalfJung`
2024-04-18 02:34:04 +00:00
Oli Scherer
126dcc618d Use less fragile error handling 2024-04-17 09:50:44 +00:00
Oli Scherer
77fe9f0a72 Validate before reporting interning errors.
validation produces much higher quality errors and already handles most of the cases
2024-04-17 09:50:44 +00:00
Oli Scherer
8b2a4f8b43 Simplify alloc id mutability check 2024-04-17 09:50:44 +00:00
Oli Scherer
140c9e10bb Deduplicate logic for checking the mutability of allocations 2024-04-17 09:50:44 +00:00
Oli Scherer
d87e9636d5 Run the "is this static mutable" logic the same way as in in_mutable_memory 2024-04-17 09:50:44 +00:00
Oli Scherer
8c9cba2be7 Validate nested static items 2024-04-17 09:50:15 +00:00
Ralf Jung
ae7b07f2dc interpret: rename base_pointer -> root_pointer
also in Miri, "base tag" -> "root tag"
2024-04-17 07:35:48 +02:00
Ralf Jung
9e239bdc76 interpret: pass MemoryKind to adjust_alloc_base_pointer 2024-04-17 07:35:48 +02:00
zetanumbers
24a24ec6ba Add simple async drop glue generation
Explainer: https://zetanumbers.github.io/book/async-drop-design.html

https://github.com/rust-lang/rust/pull/121801
2024-04-16 20:45:07 +03:00
Matthias Krüger
4971d9ffe4
Rollup merge of #124024 - RalfJung:interpret-comment, r=oli-obk
interpret: remove outdated comment

In https://github.com/rust-lang/rust/pull/107756, allocation became generally fallible, so the "only panic if there is provenance" no longer applies.

r? ``@oli-obk``
2024-04-16 17:54:46 +02:00
Ralf Jung
5b8b9cfaaa interpret: remove outdated comment 2024-04-16 17:33:12 +02:00
Ralf Jung
18bfca50f1 interpret: pass MemoryKind to before_memory_deallocation 2024-04-16 16:37:34 +02:00
Maybe Waffle
7ce867f552 Add an assertion in const eval 2024-04-16 11:56:21 +00:00