forked from OSchip/llvm-project
28 lines
1.0 KiB
LLVM
28 lines
1.0 KiB
LLVM
; RUN: llc < %s -x86-experimental-vector-shuffle-lowering=false -mattr=+avx2 | FileCheck %s
|
|
|
|
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
|
|
target triple = "x86_64-apple-darwin"
|
|
|
|
; PR21876
|
|
; The old shuffle lowering sometimes generates VZEXT nodes with both input
|
|
; and output same-sized types, here 256-bits. For instance, a v8i8 to v8i32
|
|
; zero-extend would become a (v8i32 (VZEXT v32i8)) node, which can't happen
|
|
; otherwise. The companion commit r223996 added those patterns temporarily.
|
|
; This test, along with the VR256 for AVX2 PMOVXrr instructions, should be
|
|
; removed once the old vector shuffle lowering goes away.
|
|
|
|
define void @test_avx2_pmovx_256(<8 x i8>* %tmp64, <8 x float>* %tmp75) {
|
|
; CHECK-LABEL: test_avx2_pmovx_256
|
|
; We really don't care about the generated code.
|
|
; CHECK: vpmovzxbd
|
|
; CHECK: vcvtdq2ps
|
|
; CHECK: vmovups
|
|
; CHECK: vzeroupper
|
|
; CHECK: retq
|
|
|
|
%wide.load458 = load <8 x i8>* %tmp64, align 1
|
|
%tmp68 = uitofp <8 x i8> %wide.load458 to <8 x float>
|
|
store <8 x float> %tmp68, <8 x float>* %tmp75, align 4
|
|
ret void
|
|
}
|