1 ; RUN: opt -delinearize -analyze < %s | FileCheck %s
3 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
5 ; CHECK: Inst: %tmp = load float, float* %arrayidx, align 4
6 ; CHECK-NEXT: In Loop with Header: for.inc
7 ; CHECK-NEXT: AccessFunction: {(4 * %N * %call),+,4}<nsw><%for.inc>
8 ; CHECK-NEXT: Base offset: %A
9 ; CHECK-NEXT: ArrayDecl[UnknownSize][%N] with elements of 4 bytes.
10 ; CHECK-NEXT: ArrayRef[%call][{0,+,1}<nuw><nsw><%for.inc>]
12 ; CHECK: Inst: %tmp5 = load float, float* %arrayidx4, align 4
13 ; CHECK-NEXT: In Loop with Header: for.inc
14 ; CHECK-NEXT: AccessFunction: {(4 * %call1),+,(4 * %N)}<nsw><%for.inc>
15 ; CHECK-NEXT: Base offset: %B
16 ; CHECK-NEXT: ArrayDecl[UnknownSize][%N] with elements of 4 bytes.
17 ; CHECK-NEXT: ArrayRef[{0,+,1}<nuw><nsw><%for.inc>][%call1]
19 ; Function Attrs: noinline nounwind uwtable
20 define void @mat_mul(float* %C, float* %A, float* %B, i64 %N) #0 !kernel_arg_addr_space !2 !kernel_arg_access_qual !3 !kernel_arg_type !4 !kernel_arg_base_type !4 !kernel_arg_type_qual !5 {
24 entry.split: ; preds = %entry
25 %call = tail call i64 @_Z13get_global_idj(i32 0) #3
26 %call1 = tail call i64 @_Z13get_global_idj(i32 1) #3
27 %cmp1 = icmp sgt i64 %N, 0
28 %mul = mul nsw i64 %call, %N
29 br i1 %cmp1, label %for.inc.lr.ph, label %for.end
31 for.inc.lr.ph: ; preds = %entry.split
34 for.inc: ; preds = %for.inc.lr.ph, %for.inc
35 %acc.03 = phi float [ 0.000000e+00, %for.inc.lr.ph ], [ %tmp6, %for.inc ]
36 %m.02 = phi i64 [ 0, %for.inc.lr.ph ], [ %inc, %for.inc ]
37 %add = add nsw i64 %m.02, %mul
38 %arrayidx = getelementptr inbounds float, float* %A, i64 %add
39 %tmp = load float, float* %arrayidx, align 4
40 %mul2 = mul nsw i64 %m.02, %N
41 %add3 = add nsw i64 %mul2, %call1
42 %arrayidx4 = getelementptr inbounds float, float* %B, i64 %add3
43 %tmp5 = load float, float* %arrayidx4, align 4
44 %tmp6 = tail call float @llvm.fmuladd.f32(float %tmp, float %tmp5, float %acc.03)
45 %inc = add nuw nsw i64 %m.02, 1
46 %exitcond = icmp ne i64 %inc, %N
47 br i1 %exitcond, label %for.inc, label %for.cond.for.end_crit_edge
49 for.cond.for.end_crit_edge: ; preds = %for.inc
50 %.lcssa = phi float [ %tmp6, %for.inc ]
53 for.end: ; preds = %for.cond.for.end_crit_edge, %entry.split
54 %acc.0.lcssa = phi float [ %.lcssa, %for.cond.for.end_crit_edge ], [ 0.000000e+00, %entry.split ]
55 %add7 = add nsw i64 %mul, %call1
56 %arrayidx8 = getelementptr inbounds float, float* %C, i64 %add7
57 store float %acc.0.lcssa, float* %arrayidx8, align 4
61 ; Function Attrs: nounwind readnone
62 declare i64 @_Z13get_global_idj(i32) #1
64 ; Function Attrs: nounwind readnone speculatable
65 declare float @llvm.fmuladd.f32(float, float, float) #2
67 attributes #0 = { noinline nounwind uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
68 attributes #1 = { nounwind readnone "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
69 attributes #2 = { nounwind readnone speculatable }
70 attributes #3 = { nounwind readnone }
72 !llvm.module.flags = !{!0}
75 !0 = !{i32 1, !"wchar_size", i32 4}
76 !1 = !{!"clang version 5.0.0 (trunk 303846) (llvm/trunk 303834)"}
77 !2 = !{i32 1, i32 1, i32 1, i32 0}
78 !3 = !{!"none", !"none", !"none", !"none"}
79 !4 = !{!"float*", !"float*", !"float*", !"long"}
80 !5 = !{!"", !"", !"", !""}