1 ; RUN: opt < %s -loop-vectorize -S | FileCheck %s 2 3 ; CHECK: fadd 4 ; CHECK-NEXT: fadd 5 ; CHECK-NEXT: fadd 6 ; CHECK-NEXT: fadd 7 ; CHECK-NEXT: fadd 8 ; CHECK-NEXT: fadd 9 ; CHECK-NEXT: fadd 10 ; CHECK-NEXT: fadd 11 ; CHECK-NEXT: fadd 12 ; CHECK-NEXT: fadd 13 ; CHECK-NEXT: fadd 14 ; CHECK-NEXT: fadd 15 ; CHECK-NEXT-NOT: fadd 16 17 target datalayout = "e-m:e-i64:64-n32:64" 18 target triple = "powerpc64le-ibm-linux-gnu" 19 20 define void @QLA_F3_r_veq_norm2_V(float* noalias nocapture %r, [3 x { float, float }]* noalias nocapture readonly %a, i32 signext %n) #0 { 21 entry: 22 %cmp24 = icmp sgt i32 %n, 0 23 br i1 %cmp24, label %for.cond1.preheader.preheader, label %for.end13 24 25 for.cond1.preheader.preheader: ; preds = %entry 26 br label %for.cond1.preheader 27 28 for.cond1.preheader: ; preds = %for.cond1.preheader.preheader, %for.cond1.preheader 29 %indvars.iv = phi i64 [ %indvars.iv.next, %for.cond1.preheader ], [ 0, %for.cond1.preheader.preheader ] 30 %sum.026 = phi double [ %add10.2, %for.cond1.preheader ], [ 0.000000e+00, %for.cond1.preheader.preheader ] 31 %arrayidx5.realp = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 0, i32 0 32 %arrayidx5.real = load float, float* %arrayidx5.realp, align 8 33 %arrayidx5.imagp = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 0, i32 1 34 %arrayidx5.imag = load float, float* %arrayidx5.imagp, align 8 35 %mul = fmul fast float %arrayidx5.real, %arrayidx5.real 36 %mul9 = fmul fast float %arrayidx5.imag, %arrayidx5.imag 37 %add = fadd fast float %mul9, %mul 38 %conv = fpext float %add to double 39 %add10 = fadd fast double %conv, %sum.026 40 %arrayidx5.realp.1 = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 1, i32 0 41 %arrayidx5.real.1 = load float, float* %arrayidx5.realp.1, align 8 42 %arrayidx5.imagp.1 = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 1, i32 1 43 %arrayidx5.imag.1 = load float, float* %arrayidx5.imagp.1, align 8 44 %mul.1 = fmul fast float %arrayidx5.real.1, %arrayidx5.real.1 45 %mul9.1 = fmul fast float %arrayidx5.imag.1, %arrayidx5.imag.1 46 %add.1 = fadd fast float %mul9.1, %mul.1 47 %conv.1 = fpext float %add.1 to double 48 %add10.1 = fadd fast double %conv.1, %add10 49 %arrayidx5.realp.2 = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 2, i32 0 50 %arrayidx5.real.2 = load float, float* %arrayidx5.realp.2, align 8 51 %arrayidx5.imagp.2 = getelementptr inbounds [3 x { float, float }], [3 x { float, float }]* %a, i64 %indvars.iv, i64 2, i32 1 52 %arrayidx5.imag.2 = load float, float* %arrayidx5.imagp.2, align 8 53 %mul.2 = fmul fast float %arrayidx5.real.2, %arrayidx5.real.2 54 %mul9.2 = fmul fast float %arrayidx5.imag.2, %arrayidx5.imag.2 55 %add.2 = fadd fast float %mul9.2, %mul.2 56 %conv.2 = fpext float %add.2 to double 57 %add10.2 = fadd fast double %conv.2, %add10.1 58 %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1 59 %lftr.wideiv = trunc i64 %indvars.iv.next to i32 60 %exitcond = icmp eq i32 %lftr.wideiv, %n 61 br i1 %exitcond, label %for.cond.for.end13_crit_edge, label %for.cond1.preheader 62 63 for.cond.for.end13_crit_edge: ; preds = %for.cond1.preheader 64 %add10.2.lcssa = phi double [ %add10.2, %for.cond1.preheader ] 65 %phitmp = fptrunc double %add10.2.lcssa to float 66 br label %for.end13 67 68 for.end13: ; preds = %for.cond.for.end13_crit_edge, %entry 69 %sum.0.lcssa = phi float [ %phitmp, %for.cond.for.end13_crit_edge ], [ 0.000000e+00, %entry ] 70 store float %sum.0.lcssa, float* %r, align 4 71 ret void 72 } 73 74
