1 ; RUN: opt < %s -basicaa -slp-vectorizer -dce -S -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx | FileCheck %s 2 3 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" 4 target triple = "x86_64-apple-macosx10.7.0" 5 6 @.str = private unnamed_addr constant [6 x i8] c"bingo\00", align 1 7 8 ;CHECK-LABEL: @reduce_compare( 9 ;CHECK: load <2 x double> 10 ;CHECK: fmul <2 x double> 11 ;CHECK: fmul <2 x double> 12 ;CHECK: fadd <2 x double> 13 ;CHECK: extractelement 14 ;CHECK: extractelement 15 ;CHECK: ret 16 define void @reduce_compare(double* nocapture %A, i32 %n) { 17 entry: 18 %conv = sitofp i32 %n to double 19 br label %for.body 20 21 for.body: ; preds = %for.inc, %entry 22 %indvars.iv = phi i64 [ 0, %entry ], [ %indvars.iv.next, %for.inc ] 23 %0 = shl nsw i64 %indvars.iv, 1 24 %arrayidx = getelementptr inbounds double, double* %A, i64 %0 25 %1 = load double, double* %arrayidx, align 8 26 %mul1 = fmul double %conv, %1 27 %mul2 = fmul double %mul1, 7.000000e+00 28 %add = fadd double %mul2, 5.000000e+00 29 %2 = or i64 %0, 1 30 %arrayidx6 = getelementptr inbounds double, double* %A, i64 %2 31 %3 = load double, double* %arrayidx6, align 8 32 %mul8 = fmul double %conv, %3 33 %mul9 = fmul double %mul8, 4.000000e+00 34 %add10 = fadd double %mul9, 9.000000e+00 35 %cmp11 = fcmp ogt double %add, %add10 36 br i1 %cmp11, label %if.then, label %for.inc 37 38 if.then: ; preds = %for.body 39 %call = tail call i32 (i8*, ...) @printf(i8* getelementptr inbounds ([6 x i8], [6 x i8]* @.str, i64 0, i64 0)) 40 br label %for.inc 41 42 for.inc: ; preds = %for.body, %if.then 43 %indvars.iv.next = add i64 %indvars.iv, 1 44 %lftr.wideiv = trunc i64 %indvars.iv.next to i32 45 %exitcond = icmp eq i32 %lftr.wideiv, 100 46 br i1 %exitcond, label %for.end, label %for.body 47 48 for.end: ; preds = %for.inc 49 ret void 50 } 51 52 declare i32 @printf(i8* nocapture, ...) 53 54
