summary refs log tree commit diff
path: root/compiler/rustc_codegen_ssa/src/mir
AgeCommit message (Collapse)AuthorLines
2022-12-04Auto merge of #104535 - mikebenfield:discr-fix, r=pnkfelixbors-9/+18
rustc_codegen_ssa: Fix for codegen_get_discr When doing the optimized implementation of getting the discriminant, the arithmetic needs to be done in the tag type so wrapping behavior works correctly. Fixes #104519
2022-11-27Prefer doc comments over `//`-comments in compilerMaybe Waffle-2/+2
2022-11-24Properly handle `Pin<&mut dyn* Trait>` receiver in codegenMichael Goulet-6/+24
2022-11-19deduplicate constant evaluation in cranelift backendRalf Jung-1/+8
also sync LLVM and cranelift structure a bit
2022-11-18rustc_codegen_ssa: Fix for codegen_get_discrMichael Benfield-9/+18
When doing the optimized implementation of getting the discriminant, the arithmetic needs to be done in the tag type so wrapping behavior works correctly. Fixes #104519
2022-11-17Auto merge of #103138 - nnethercote:merge-BBs, r=bjorn3bors-331/+406
Merge basic blocks where possible when generating LLVM IR. r? `@ghost`
2022-11-16cleanup and dedupe CTFE and Miri error reportingRalf Jung-1/+1
2022-11-16Merge basic blocks where possible when generating LLVM IR.Nicholas Nethercote-135/+251
In `codegen_assert_terminator` we decide if a BB's successor is a candidate for merging, which requires that it be the only successor, and that it only have one predecessor. That result then gets passed down, and if it reaches `funclet_br` with the appropriate BB characteristics, then no `br` instruction is issued, a `MergingSucc::True` result is passed back, and the merging proceeds in `codegen_block`. The commit also adds `CachedLlbb`, a new type to help keep track of each BB that has been merged into its predecessor.
2022-11-16Use `&mut Bx` more.Nicholas Nethercote-206/+165
For the next commit, `FunctionCx::codegen_*_terminator` need to take a `&mut Bx` instead of consuming a `Bx`. This triggers a cascade of similar changes across multiple functions. The resulting code is more concise and replaces many `&mut bx` expressions with `bx`.
2022-11-15Introduce composite debuginfo.Camille GILLOT-7/+64
2022-11-13add is_sized method on Abi and Layout, and use itRalf Jung-4/+4
2022-11-11rustc_codegen_ssa: Better code generation for niche discriminants.Michael Benfield-49/+145
In some cases we can avoid arithmetic before checking whether a niche represents an untagged variant. This is relevant to #101872
2022-10-31Use `br` instead of `switch` in more cases.Nicholas Nethercote-2/+28
`codegen_switchint_terminator` already uses `br` instead of `switch` when there is one normal target plus the `otherwise` target. But there's another common case with two normal targets and an `otherwise` target that points to an empty unreachable BB. This comes up a lot when switching on the tags of enums that use niches. The pattern looks like this: ``` bb1: ; preds = %bb6 %3 = load i8, ptr %_2, align 1, !range !9, !noundef !4 %4 = sub i8 %3, 2 %5 = icmp eq i8 %4, 0 %_6 = select i1 %5, i64 0, i64 1 switch i64 %_6, label %bb3 [ i64 0, label %bb4 i64 1, label %bb2 ] bb3: ; preds = %bb1 unreachable ``` This commit adds code to convert the `switch` to a `br`: ``` bb1: ; preds = %bb6 %3 = load i8, ptr %_2, align 1, !range !9, !noundef !4 %4 = sub i8 %3, 2 %5 = icmp eq i8 %4, 0 %_6 = select i1 %5, i64 0, i64 1 %6 = icmp eq i64 %_6, 0 br i1 %6, label %bb4, label %bb2 bb3: ; No predecessors! unreachable ``` This has a surprisingly large effect on compile times, with reductions of 5% on debug builds of some crates. The reduction is all due to LLVM taking less time. Maybe LLVM is just much better at handling `br` than `switch`. The resulting code is still suboptimal. - The `icmp`, `select`, `icmp` sequence is silly, converting an `i1` to an `i64` and back to an `i1`. But with the current code structure it's hard to avoid, and LLVM will easily clean it up, in opt builds at least. - `bb3` is usually now truly dead code (though not always, so it can't be removed universally).
2022-10-25Clarify some cleanup stuff.Nicholas Nethercote-6/+10
- Rearrange the match in `llbb_with_landing_pad` so the `(Some,Some)` cases are together. - Add assertions to indicate two MSVC-only paths.
2022-10-25Rename two `TerminatorCodegenHelper` methods.Nicholas Nethercote-13/+16
`TerminatorCodegenHelper` has three methods `llblock`, `llbb`, and `lltarget`. They're all similar, but the names given no indication of the differences. This commit renames `lltarget` as `llbb_with_landing_pad`, and `llblock` as `llbb_with_cleanup`. These aren't fantastic names, but at least it's now clear that `llbb` is the lowest-level of the three and the other two wrap it.
2022-10-25rustc_codegen_ssa: use more consistent naming.Nicholas Nethercote-30/+31
Ensure: - builders always have a `bx` suffix; - backend basic blocks always have an `llbb` suffix, - paired builders and basic blocks have consistent prefixes.
2022-10-14Allow dyn* upcastingMichael Goulet-21/+6
2022-10-13Rollup merge of #102641 - eholk:dyn-star-box, r=compiler-errorsYuki Okushi-0/+9
Support casting boxes to dyn* Boxes have a pointer type at codegen time which LLVM does not allow to be transparently converted to an integer. Work around this by inserting a `ptrtoint` instruction if the argument is a pointer. r? ``@compiler-errors`` Fixes #102427
2022-10-12Add a fixmeEric Holk-0/+3
2022-10-07Cast vtable type tooEric Holk-0/+1
2022-10-06Remove `mir::CastKind::Misc`ouz-a-2/+8
2022-10-04Support casting boxes to dyn*Eric Holk-0/+5
2022-10-02Remove type argument of array_alloca and rename to byte_array_allocabjorn3-1/+1
2022-10-01Merge apply_attrs_callsite into call and invokebjorn3-7/+11
Some codegen backends are not able to apply callsite attrs after the fact.
2022-09-20Auto merge of #99806 - oli-obk:unconstrained_opaque_type, r=estebankbors-0/+16
Allow patterns to constrain the hidden type of opaque types fixes #96572 reverts a revert as original PR was a perf regression that was fixed by reverting it: https://github.com/rust-lang/rust/pull/99368#issuecomment-1186587864) TODO: * check if https://github.com/rust-lang/rust/issues/99685 is avoided
2022-09-17Auto merge of #98588 - b-naber:valtrees-cleanup, r=lcnrbors-18/+18
Use only ty::Unevaluated<'tcx, ()> in type system r? `@lcnr`
2022-09-16Revert "Revert "Rollup merge of #98582 - oli-obk:unconstrained_opaque_type, ↵Oli Scherer-0/+16
r=estebank"" This reverts commit 4a742a691e7dd2522bad68b86fe2fd5a199d5561.
2022-09-14Auto merge of #101212 - eholk:dyn-star, r=compiler-errorsbors-2/+103
Initial implementation of dyn* This PR adds extremely basic and incomplete support for [dyn*](https://smallcultfollowing.com/babysteps//blog/2022/03/29/dyn-can-we-make-dyn-sized/). The goal is to get something in tree behind a flag to make collaboration easier, and also to make sure the implementation so far is not unreasonable. This PR does quite a few things: * Introduce `dyn_star` feature flag * Adds parsing for `dyn* Trait` types * Defines `dyn* Trait` as a sized type * Adds support for explicit casts, like `42usize as dyn* Debug` * Including const evaluation of such casts * Adds codegen for drop glue so things are cleaned up properly when a `dyn* Trait` object goes out of scope * Adds codegen for method calls, at least for methods that take `&self` Quite a bit is still missing, but this gives us a starting point. Note that this is never intended to become stable surface syntax for Rust, but rather `dyn*` is planned to be used as an implementation detail for async functions in dyn traits. Joint work with `@nikomatsakis` and `@compiler-errors.` r? `@bjorn3`
2022-09-13Address code review commentsEric Holk-1/+1
2022-09-13use ty::Unevaluated<'tcx, ()> in type systemb-naber-18/+18
2022-09-13Auto merge of #99556 - davidtwco:collapse-debuginfo, r=wesleywiserbors-4/+4
ssa: implement `#[collapse_debuginfo]` cc #39153 rust-lang/compiler-team#386 Debuginfo line information for macro invocations are collapsed by default - line information are replaced by the line of the outermost expansion site. Using `-Zdebug-macros` disables this behaviour. When the `collapse_debuginfo` feature is enabled, the default behaviour is reversed so that debuginfo is not collapsed by default. In addition, the `#[collapse_debuginfo]` attribute is available and can be applied to macro definitions which will then have their line information collapsed. r? rust-lang/wg-debugging
2022-09-12Make dyn-trait-method workMichael Goulet-1/+24
2022-09-12Rename some variantsMichael Goulet-4/+4
2022-09-12Construct dyn* during const interpMichael Goulet-1/+1
2022-09-12Use principal of cast target as dyn-star trait ref in codegenMichael Goulet-7/+6
2022-09-12Call destructors when dyn* object goes out of scopeEric Holk-5/+32
2022-09-12dyn* through more typechecking and MIREric Holk-2/+54
2022-09-10Auto merge of #101483 - oli-obk:guaranteed_opt, r=fee1-deadbors-6/+2
The `<*const T>::guaranteed_*` methods now return an option for the unknown case cc https://github.com/rust-lang/rust/issues/53020#issuecomment-1236932443 I chose `0` for "not equal" and `1` for "equal" and left `2` for the unknown case so backends can just forward to raw pointer equality and it works ✨ r? `@fee1-dead` or `@lcnr` cc `@rust-lang/wg-const-eval`
2022-09-09The `<*const T>::guaranteed_*` methods now return an option for the unknown caseOli Scherer-6/+2
2022-09-09Use memset when repeating 128bit zero valueTomasz Miąsko-1/+1
2022-09-07Change name of "dataful" variant to "untagged"Michael Benfield-4/+4
This is in anticipation of a new enum layout, in which the niche optimization may be applied even when multiple variants have data.
2022-09-07ssa: implement `#[collapse_debuginfo]`David Wood-4/+4
Debuginfo line information for macro invocations are collapsed by default - line information are replaced by the line of the outermost expansion site. Using `-Zdebug-macros` disables this behaviour. When the `collapse_debuginfo` feature is enabled, the default behaviour is reversed so that debuginfo is not collapsed by default. In addition, the `#[collapse_debuginfo]` attribute is available and can be applied to macro definitions which will then have their line information collapsed. Signed-off-by: David Wood <david.wood@huawei.com>
2022-09-07Auto merge of #98332 - oli-obk:assume, r=wesleywiserbors-9/+9
Lower the assume intrinsic to a MIR statement This makes https://github.com/rust-lang/rust/pull/96862#issuecomment-1153739068 easier and will generally allow us to cheaply insert assume intrinsic calls in mir building. r? rust-lang/wg-mir-opt
2022-09-06Generalize the Assume intrinsic statement to a general Intrinsic statementOli Scherer-10/+9
2022-09-06Lower the assume intrinsic to a MIR statementOli Scherer-4/+5
2022-09-06Remove dead broken code from const zst handling in backendsOli Scherer-8/+1
2022-09-04Make `const_eval_select` a real intrinsicDeadbeef-2/+1
2022-09-01Auto merge of #100958 - mikebenfield:workaround, r=nikicbors-11/+0
compiler/rustc_codegen_ssa/src/mir/place.rs: Remove LLVM bug workaround This memset was inserted as a workaround to Rust issue #34427, which was an LLVM bug that apparently no longer manifests.
2022-08-31Fix a bunch of typoDezhi Wu-1/+1
This PR will fix some typos detected by [typos]. I only picked the ones I was sure were spelling errors to fix, mostly in the comments. [typos]: https://github.com/crate-ci/typos
2022-08-29Rollup merge of #99027 - tmiasko:basic-blocks, r=oli-obkMatthias Krüger-7/+7
Replace `Body::basic_blocks()` with field access Since the refactoring in #98930, it is possible to borrow the basic blocks independently from other parts of MIR by accessing the `basic_blocks` field directly. Replace unnecessary `Body::basic_blocks()` method with a direct field access, which has an additional benefit of borrowing the basic blocks only.
generated by cgit-pink 1.4.1-3-g733a5 (git 2.37.1) at 2026-07-03 20:31:04 +0000