Home | History | Annotate | Download | only in X86
      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-macosx10.8.0"
      5 
      6 ; At this point we can't vectorize only parts of the tree.
      7 
      8 ; CHECK: test
      9 ; CHECK: insertelement <2 x i8>
     10 ; CHECK: insertelement <2 x i8>
     11 ; CHECK: sitofp <2 x i8>
     12 ; CHECK: fmul <2 x double>
     13 ; CHECK: ret
     14 define i32 @test(double* nocapture %A, i8* nocapture %B) {
     15 entry:
     16   %0 = load i8* %B, align 1
     17   %arrayidx1 = getelementptr inbounds i8* %B, i64 1
     18   %1 = load i8* %arrayidx1, align 1
     19   %add = add i8 %0, 3
     20   %add4 = add i8 %1, 3
     21   %conv6 = sitofp i8 %add to double
     22   %conv7 = sitofp i8 %add4 to double 
     23   %mul = fmul double %conv6, %conv6
     24   %add8 = fadd double %mul, 1.000000e+00
     25   %mul9 = fmul double %conv7, %conv7
     26   %add10 = fadd double %mul9, 1.000000e+00
     27   %mul11 = fmul double %add8, %add8
     28   %add12 = fadd double %mul11, 1.000000e+00
     29   %mul13 = fmul double %add10, %add10
     30   %add14 = fadd double %mul13, 1.000000e+00
     31   %mul15 = fmul double %add12, %add12
     32   %add16 = fadd double %mul15, 1.000000e+00
     33   %mul17 = fmul double %add14, %add14
     34   %add18 = fadd double %mul17, 1.000000e+00
     35   %mul19 = fmul double %add16, %add16
     36   %add20 = fadd double %mul19, 1.000000e+00
     37   %mul21 = fmul double %add18, %add18
     38   %add22 = fadd double %mul21, 1.000000e+00
     39   %mul23 = fmul double %add20, %add20
     40   %add24 = fadd double %mul23, 1.000000e+00
     41   %mul25 = fmul double %add22, %add22
     42   %add26 = fadd double %mul25, 1.000000e+00
     43   store double %add24, double* %A, align 8
     44   %arrayidx28 = getelementptr inbounds double* %A, i64 1
     45   store double %add26, double* %arrayidx28, align 8
     46   ret i32 undef
     47 }
     48