1 ; RUN: llc < %s -march=x86 -mattr=+sse,+sse2,+sse41 | FileCheck %s 2 3 target datalayout = "e-p:32:32" 4 target triple = "i686-apple-darwin8.7.2" 5 6 define i16 @test1(float %f) nounwind { 7 %tmp = insertelement <4 x float> undef, float %f, i32 0 ; <<4 x float>> [#uses=1] 8 %tmp10 = insertelement <4 x float> %tmp, float 0.000000e+00, i32 1 ; <<4 x float>> [#uses=1] 9 %tmp11 = insertelement <4 x float> %tmp10, float 0.000000e+00, i32 2 ; <<4 x float>> [#uses=1] 10 %tmp12 = insertelement <4 x float> %tmp11, float 0.000000e+00, i32 3 ; <<4 x float>> [#uses=1] 11 %tmp28 = tail call <4 x float> @llvm.x86.sse.sub.ss( <4 x float> %tmp12, <4 x float> < float 1.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > ) ; <<4 x float>> [#uses=1] 12 %tmp37 = tail call <4 x float> @llvm.x86.sse.mul.ss( <4 x float> %tmp28, <4 x float> < float 5.000000e-01, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > ) ; <<4 x float>> [#uses=1] 13 %tmp48 = tail call <4 x float> @llvm.x86.sse.min.ss( <4 x float> %tmp37, <4 x float> < float 6.553500e+04, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00 > ) ; <<4 x float>> [#uses=1] 14 %tmp59 = tail call <4 x float> @llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> zeroinitializer ) ; <<4 x float>> [#uses=1] 15 %tmp.upgrd.1 = tail call i32 @llvm.x86.sse.cvttss2si( <4 x float> %tmp59 ) ; <i32> [#uses=1] 16 %tmp69 = trunc i32 %tmp.upgrd.1 to i16 ; <i16> [#uses=1] 17 ret i16 %tmp69 18 ; CHECK: test1: 19 ; CHECK: subss LCPI0_ 20 ; CHECK: mulss LCPI0_ 21 ; CHECK: minss LCPI0_ 22 } 23 24 define i16 @test2(float %f) nounwind { 25 %tmp28 = fsub float %f, 1.000000e+00 ; <float> [#uses=1] 26 %tmp37 = fmul float %tmp28, 5.000000e-01 ; <float> [#uses=1] 27 %tmp375 = insertelement <4 x float> undef, float %tmp37, i32 0 ; <<4 x float>> [#uses=1] 28 %tmp48 = tail call <4 x float> @llvm.x86.sse.min.ss( <4 x float> %tmp375, <4 x float> < float 6.553500e+04, float undef, float undef, float undef > ) ; <<4 x float>> [#uses=1] 29 %tmp59 = tail call <4 x float> @llvm.x86.sse.max.ss( <4 x float> %tmp48, <4 x float> < float 0.000000e+00, float undef, float undef, float undef > ) ; <<4 x float>> [#uses=1] 30 %tmp = tail call i32 @llvm.x86.sse.cvttss2si( <4 x float> %tmp59 ) ; <i32> [#uses=1] 31 %tmp69 = trunc i32 %tmp to i16 ; <i16> [#uses=1] 32 ret i16 %tmp69 33 ; CHECK: test2: 34 ; CHECK: addss LCPI1_ 35 ; CHECK: mulss LCPI1_ 36 ; CHECK: minss LCPI1_ 37 } 38 39 declare <4 x float> @llvm.x86.sse.sub.ss(<4 x float>, <4 x float>) 40 41 declare <4 x float> @llvm.x86.sse.mul.ss(<4 x float>, <4 x float>) 42 43 declare <4 x float> @llvm.x86.sse.min.ss(<4 x float>, <4 x float>) 44 45 declare <4 x float> @llvm.x86.sse.max.ss(<4 x float>, <4 x float>) 46 47 declare i32 @llvm.x86.sse.cvttss2si(<4 x float>) 48 49 50 declare <4 x float> @llvm.x86.sse41.round.ss(<4 x float>, <4 x float>, i32) 51 declare <4 x float> @f() 52 53 define <4 x float> @test3(<4 x float> %A, float *%b, i32 %C) nounwind { 54 %a = load float *%b 55 %B = insertelement <4 x float> undef, float %a, i32 0 56 %X = call <4 x float> @llvm.x86.sse41.round.ss(<4 x float> %A, <4 x float> %B, i32 4) 57 ret <4 x float> %X 58 ; CHECK: test3: 59 ; CHECK: roundss $4, (%eax), %xmm0 60 } 61 62 define <4 x float> @test4(<4 x float> %A, float *%b, i32 %C) nounwind { 63 %a = load float *%b 64 %B = insertelement <4 x float> undef, float %a, i32 0 65 %q = call <4 x float> @f() 66 %X = call <4 x float> @llvm.x86.sse41.round.ss(<4 x float> %q, <4 x float> %B, i32 4) 67 ret <4 x float> %X 68 ; CHECK: test4: 69 ; CHECK: movss (%eax), %xmm 70 ; CHECK: call 71 ; CHECK: roundss $4, %xmm{{.*}}, %xmm0 72 } 73
