about summary refs log tree commit diff
path: root/compiler/rustc_codegen_gcc/src/intrinsic/simd.rs
AgeCommit message (Collapse)AuthorLines
2025-08-26Merge commit 'feb42827f11a7ae241ceecc81e9ae556fb6ba214' into ↵Guillaume Gomez-1/+0
subtree-update_cg_gcc_2025-08-26
2025-08-04Merge commit '482e8540a1b757ed7bccc2041c5400f051fdb01e' into ↵Guillaume Gomez-0/+81
subtree-update_cg_gcc_2025-08-04
2025-07-07compiler: Parse `p-` specs in datalayout string, allow definition of custom ↵Edoardo Marangoni-1/+1
default data address space
2025-06-29Rollup merge of #142078 - sayantn:more-intrinsics, r=workingjubileeGuillaume Gomez-0/+2
Add SIMD funnel shift and round-to-even intrinsics This PR adds 3 new SIMD intrinsics - `simd_funnel_shl` - funnel shift left - `simd_funnel_shr` - funnel shift right - `simd_round_ties_even` (vector version of `round_ties_even_fN`) TODO (future PR): implement `simd_fsh{l,r}` in miri, cg_gcc and cg_clif (it is surprisingly hard to implement without branches, the common tricks that rotate uses doesn't work because we have 2 elements now. e.g, the `-n&31` trick used by cg_gcc to implement rotate doesn't work with this because then `fshl(a, b, 0)` will be `a | b`) [#t-compiler > More SIMD intrinsics](https://rust-lang.zulipchat.com/#narrow/channel/131828-t-compiler/topic/More.20SIMD.20intrinsics/with/522130286) `@rustbot` label T-compiler T-libs A-intrinsics F-core_intrinsics r? `@workingjubilee`
2025-06-27rustc_codegen_gcc: Fix clippy::manual_is_multiple_ofPhilipp Krones-2/+3
2025-06-18Merge commit 'fda0bb9588912a3e0606e880ca9f6e913cf8a5a4' into ↵Guillaume Gomez-1/+3
subtree-update_cg_gcc_2025-06-18
2025-06-15Implement `simd_round_ties_even` for miri, cg_clif and cg_gccsayantn-0/+2
2025-05-30Use layout field of OperandRef in generic_simd_intrinsicbjorn3-46/+41
2025-04-20Rollup merge of #137953 - RalfJung:simd-intrinsic-masks, r=WaffleLapkinChris Denton-14/+11
simd intrinsics with mask: accept unsigned integer masks, and fix some of the errors It's not clear at all why the mask would have to be signed, it is anyway interpreted bitwise. The backend should just make sure that works no matter the surface-level type; our LLVM backend already does this correctly. The note of "the mask may be widened, which only has the correct behavior for signed integers" explains... nothing? Why can't the code do the widening correctly? If necessary, just cast to the signed type first... Also while we are at it, fix the errors. For simd_masked_load/store, the errors talked about the "third argument" but they meant the first argument (the mask is the first argument there). They also used the wrong type for `expected_element`. I have extremely low confidence in the GCC part of this PR. See [discussion on Zulip](https://rust-lang.zulipchat.com/#narrow/channel/257879-project-portable-simd/topic/On.20the.20sign.20of.20masks)
2025-04-20simd intrinsics with mask: accept unsigned integer masksRalf Jung-14/+11
2025-04-18Merge commit 'db1a31c243a649e1fe20f5466ba181da5be35c14' into ↵Guillaume Gomez-3/+7
subtree-update_cg_gcc_2025-04-18
2025-02-25remove `simd_fpow` and `simd_fpowi`Folkert de Vries-20/+8
2025-02-08Rustfmtbjorn3-229/+237
2025-02-04cg_gcc: Directly use rustc_abi instead of reexportsJubilee Young-1/+1
2025-01-13Merge commit '59a81c2ca1edc88ad3ac4b27a8e03977ffb8e73a' into ↵Antoni Boucher-1/+2
subtree-update_cg_gcc_2025_01_12
2024-11-23Add simd_relaxed_fma intrinsicCaleb Zulawski-0/+1
2024-11-18use `TypingEnv` when no `infcx` is availablelcnr-4/+6
the behavior of the type system not only depends on the current assumptions, but also the currentnphase of the compiler. This is mostly necessary as we need to decide whether and how to reveal opaque types. We track this via the `TypingMode`.
2024-10-19Fix testsMichael Goulet-4/+8
2024-10-04Use wide pointers consistenly across the compilerUrgau-2/+2
2024-09-27Merge commit '3187d32079b817522cc17413ec9185b130daf693' into subtree-updateGuillaume Gomez-15/+40
2024-09-22Reformat using the new identifier sorting from rustfmtMichael Goulet-238/+230
2024-09-17Rename supertraits of `CodegenMethods`.Nicholas Nethercote-1/+1
Supertraits of `BuilderMethods` are all called `XyzBuilderMethods`. Supertraits of `CodegenMethods` are all called `XyzMethods`. This commit changes the latter to `XyzCodegenMethods`, for consistency.
2024-09-14simd_shuffle: require index argument to be a vectorRalf Jung-18/+7
2024-08-13simd_shuffle intrinsic: allow argument to be passed as vector (not just as ↵Ralf Jung-7/+12
array)
2024-07-29Reformat `use` declarations.Nicholas Nethercote-3/+1
The previous commit updated `rustfmt.toml` appropriately. This commit is the outcome of running `x fmt --all` with the new formatting options.
2024-07-10Merge commit '98ed962c7d3eebe12c97588e61245273d265e72f' into masterGuillaume Gomez-29/+166
2024-05-22Stop using `to_hir_binop` in codegenScott McMurray-6/+7
2024-04-24Auto merge of #122053 - erikdesjardins:alloca, r=nikicbors-2/+2
Stop using LLVM struct types for alloca The alloca type has no semantic meaning, only the size (and alignment, but we specify it explicitly) matter. Using `[N x i8]` is a more direct way to specify that we want `N` bytes, and avoids relying on LLVM's struct layout. It is likely that a future LLVM version will change to an untyped alloca representation. Split out from #121577. r? `@ghost`
2024-04-11use [N x i8] for alloca typesErik Desjardins-2/+2
2024-04-11Make `PlaceRef` hold a `PlaceValue` for the non-layout fields (like ↵Scott McMurray-1/+1
`OperandRef` does)
2024-03-22Make RawPtr take Ty and Mutbl separatelyMichael Goulet-12/+12
2024-03-05Merge commit 'b385428e3ddf330805241e7758e773f933357c4b' into ↵Guillaume Gomez-24/+206
subtree-update_cg_gcc_2024-03-05
2024-02-21remove simd_reduce_{min,max}_nanlessRalf Jung-3/+0
2024-02-21make simd_reduce_{mul,add}_unordered use only the 'reassoc' flag, not all ↵Ralf Jung-2/+2
fast-math flags
2023-12-24Remove `Session` methods that duplicate `DiagCtxt` methods.Nicholas Nethercote-2/+2
Also add some `dcx` methods to types that wrap `TyCtxt`, for easier access.
2023-08-03Forbid old-style `simd_shuffleN` intrinsicsOli Scherer-195/+250
2023-07-20UPDATE - replace expected_simd error with one from codegen_ssaJhonny Bill Mena-27/+15
Here I am assuming we want to treat these parameters (input, first, second, third, return) as translatable
2023-07-19UPDATE - replace gcc monomorphization errors with ssa onesJhonny Bill Mena-39/+40
Reduces error duplication and makes it more consistent across backends
2023-06-19Merge commit '1bbee3e217d75e7bc3bfe5d8c1b35e776fce96e6' into ↵Antoni Boucher-5/+13
sync-cg_gcc-2023-06-19
2023-03-05Merge commit '08a6d6e16b5efe217123e780398969946266268f' into ↵Antoni Boucher-257/+582
sync-cg_gcc-2023-03-04
2023-02-14s/eval_usize/eval_target_usize/ for clarityOli Scherer-174/+233
2022-10-01Merge apply_attrs_callsite into call and invokebjorn3-1/+1
Some codegen backends are not able to apply callsite attrs after the fact.
2022-09-24remove commentEllis Hoag-1/+0
2022-09-24Add monomorphization errorsEllis Hoag-142/+61
2022-06-07Remove unused macro ruleAntoni Boucher-3/+0
2022-06-06Merge commit 'e8dca3e87d164d2806098c462c6ce41301341f68' into sync_from_cg_gccAntoni Boucher-6/+571
2022-05-27Finish bumping stage0Mark Rousskov-0/+1
It looks like the last time had left some remaining cfg's -- which made me think that the stage0 bump was actually successful. This brings us to a released 1.62 beta though.
2022-03-26Merge commit '39683d8eb7a32a74bea96ecbf1e87675d3338506' into ↵bjorn3-0/+21
sync_cg_gcc-2022-03-26
2021-12-15Remove unnecessary sigils around `Symbol::as_str()` calls.Nicholas Nethercote-1/+1
2021-08-15Merge commit 'e228f0c16ea8c34794a6285bf57aab627c26b147' into libgccjit-codegenAntoni Boucher-835/+1
generated by cgit-pink 1.4.1-3-g733a5 (git 2.37.1) at 2025-11-07 07:01:32 +0000