mikros/rust
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Auto merge of #126163 - RalfJung:simd-packed, r=calebzulawski,workingjubilee

Browse Source 
simd packed types: remove outdated comment, extend codegen test

It seems like https://github.com/rust-lang/rust/pull/125311 made that check in codegen unnecessary?

r? `@workingjubilee` `@calebzulawski`
This commit is contained in:
bors 2024-06-09 07:28:47 +00:00
commit b3ca6ee18a
2 changed files with 25 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
compiler/rustc_codegen_llvm/src/intrinsic.rs
View File

@ -1109,10 +1109,12 @@ macro_rules! require_simd {
tcx.normalize_erasing_late_bound_regions(ty::ParamEnv::reveal_all(), callee_ty.fn_sig(tcx)); tcx.normalize_erasing_late_bound_regions(ty::ParamEnv::reveal_all(), callee_ty.fn_sig(tcx));
let arg_tys = sig.inputs(); let arg_tys = sig.inputs();
// Vectors must be immediates (non-power-of-2 #[repr(packed)] are not) // Sanity-check: all vector arguments must be immediates.
for (ty, arg) in arg_tys.iter().zip(args) { if cfg!(debug_assertions) {
if ty.is_simd() && !matches!(arg.val, OperandValue::Immediate(_)) { for (ty, arg) in arg_tys.iter().zip(args) {
return_error!(InvalidMonomorphization::SimdArgument { span, name, ty: *ty }); if ty.is_simd() {
assert!(matches!(arg.val, OperandValue::Immediate(_)));
}
} }
} }

23
tests/codegen/simd/packed-simd.rs
View File

@ -9,10 +9,11 @@
use core::{mem, ptr}; use core::{mem, ptr};
// Test codegen for not only "packed" but also "fully aligned" SIMD types, and conversion between // Test codegen for not only "packed" but also "fully aligned" SIMD types, and conversion between
// A repr(packed,simd) type with 3 elements can't exceed its element alignment, // them. A repr(packed,simd) type with 3 elements can't exceed its element alignment, whereas the
// whereas the same type as repr(simd) will instead have padding. // same type as repr(simd) will instead have padding.
#[repr(simd, packed)] #[repr(simd, packed)]
#[derive(Copy, Clone)]
pub struct Simd<T, const N: usize>([T; N]); pub struct Simd<T, const N: usize>([T; N]);
#[repr(simd)] #[repr(simd)]
@ -28,11 +29,11 @@ fn load<T, const N: usize>(v: Simd<T, N>) -> FullSimd<T, N> {
} }
} }
// CHECK-LABEL: square_packed // CHECK-LABEL: square_packed_full
// CHECK-SAME: ptr{{[a-z_ ]*}} sret([[RET_TYPE:[^)]+]]) [[RET_ALIGN:align (8|16)]]{{[^%]*}} [[RET_VREG:%[_0-9]*]] // CHECK-SAME: ptr{{[a-z_ ]*}} sret([[RET_TYPE:[^)]+]]) [[RET_ALIGN:align (8|16)]]{{[^%]*}} [[RET_VREG:%[_0-9]*]]
// CHECK-SAME: ptr{{[a-z_ ]*}} align 4 // CHECK-SAME: ptr{{[a-z_ ]*}} align 4
#[no_mangle] #[no_mangle]
pub fn square_packed(x: Simd<f32, 3>) -> FullSimd<f32, 3> { pub fn square_packed_full(x: Simd<f32, 3>) -> FullSimd<f32, 3> {
// CHECK-NEXT: start // CHECK-NEXT: start
// noopt: alloca [[RET_TYPE]], [[RET_ALIGN]] // noopt: alloca [[RET_TYPE]], [[RET_ALIGN]]
// CHECK: load <3 x float> // CHECK: load <3 x float>
@ -42,3 +43,17 @@ pub fn square_packed(x: Simd<f32, 3>) -> FullSimd<f32, 3> {
// CHECK-NEXT: ret void // CHECK-NEXT: ret void
unsafe { intrinsics::simd_mul(x, x) } unsafe { intrinsics::simd_mul(x, x) }
} }
// CHECK-LABEL: square_packed
// CHECK-SAME: ptr{{[a-z_ ]*}} sret([[RET_TYPE:[^)]+]]) [[RET_ALIGN:align 4]]{{[^%]*}} [[RET_VREG:%[_0-9]*]]
// CHECK-SAME: ptr{{[a-z_ ]*}} align 4
#[no_mangle]
pub fn square_packed(x: Simd<f32, 3>) -> Simd<f32, 3> {
// CHECK-NEXT: start
// CHECK-NEXT: load <3 x float>
// noopt-NEXT: load <3 x float>
// CHECK-NEXT: [[VREG:%[a-z0-9_]+]] = fmul <3 x float>
// CHECK-NEXT: store <3 x float> [[VREG]], ptr [[RET_VREG]], [[RET_ALIGN]]
// CHECK-NEXT: ret void
unsafe { intrinsics::simd_mul(x, x) }
}