1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc -mtriple=x86_64-unknown-unknown -mattr=sse2 < %s | FileCheck --check-prefix=SSE %s
3 ; RUN: llc -mtriple=x86_64-unknown-unknown -mattr=sse4.1 < %s | FileCheck --check-prefix=SSE %s
4 ; RUN: llc -mtriple=x86_64-unknown-unknown -mattr=avx < %s | FileCheck --check-prefix=AVX %s
6 ; PR21507 - https://llvm.org/bugs/show_bug.cgi?id=21507
7 ; Each function should be a single math op; no extra moves.
10 define <4 x float> @recip(<4 x float> %x) {
13 ; SSE-NEXT: rcpss %xmm0, %xmm0
18 ; AVX-NEXT: vrcpss %xmm0, %xmm0, %xmm0
20 %y = tail call <4 x float> @llvm.x86.sse.rcp.ss(<4 x float> %x)
21 %shuf = shufflevector <4 x float> %y, <4 x float> %x, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
25 define <4 x float> @recip_square_root(<4 x float> %x) {
26 ; SSE-LABEL: recip_square_root:
28 ; SSE-NEXT: rsqrtss %xmm0, %xmm0
31 ; AVX-LABEL: recip_square_root:
33 ; AVX-NEXT: vrsqrtss %xmm0, %xmm0, %xmm0
35 %y = tail call <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float> %x)
36 %shuf = shufflevector <4 x float> %y, <4 x float> %x, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
40 define <4 x float> @square_root(<4 x float> %x) {
41 ; SSE-LABEL: square_root:
43 ; SSE-NEXT: sqrtss %xmm0, %xmm0
46 ; AVX-LABEL: square_root:
48 ; AVX-NEXT: vsqrtss %xmm0, %xmm0, %xmm0
50 %y = tail call <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float> %x)
51 %shuf = shufflevector <4 x float> %y, <4 x float> %x, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
55 define <2 x double> @square_root_double(<2 x double> %x) {
56 ; SSE-LABEL: square_root_double:
58 ; SSE-NEXT: sqrtsd %xmm0, %xmm0
61 ; AVX-LABEL: square_root_double:
63 ; AVX-NEXT: vsqrtsd %xmm0, %xmm0, %xmm0
65 %y = tail call <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double> %x)
66 %shuf = shufflevector <2 x double> %y, <2 x double> %x, <2 x i32> <i32 0, i32 3>
67 ret <2 x double> %shuf
70 declare <4 x float> @llvm.x86.sse.rcp.ss(<4 x float>)
71 declare <4 x float> @llvm.x86.sse.rsqrt.ss(<4 x float>)
72 declare <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float>)
73 declare <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double>)