1; RUN: opt -S %loadPolly -basic-aa -polly-pattern-matching-based-opts=false -polly-vectorizer=polly                         -polly-opt-isl -polly-print-ast -disable-output < %s | FileCheck %s
2; RUN: opt -S %loadPolly -basic-aa -polly-pattern-matching-based-opts=false -polly-vectorizer=stripmine                     -polly-opt-isl -polly-print-ast -disable-output < %s | FileCheck %s
3; RUN: opt -S %loadPolly -basic-aa -polly-pattern-matching-based-opts=false -polly-vectorizer=polly -polly-prevect-width=16 -polly-opt-isl -polly-print-ast -disable-output < %s | FileCheck %s -check-prefix=VEC16
4
5target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
6
7@C = common global [1536 x [1536 x float]] zeroinitializer, align 16
8@A = common global [1536 x [1536 x float]] zeroinitializer, align 16
9@B = common global [1536 x [1536 x float]] zeroinitializer, align 16
10
11; Function Attrs: nounwind uwtable
12define void @foo() #0 {
13entry:
14  br label %entry.split
15
16entry.split:                                      ; preds = %entry
17  br label %for.cond1.preheader
18
19for.cond1.preheader:                              ; preds = %entry.split, %for.inc28
20  %indvar4 = phi i64 [ 0, %entry.split ], [ %indvar.next5, %for.inc28 ]
21  br label %for.body3
22
23for.body3:                                        ; preds = %for.cond1.preheader, %for.inc25
24  %indvar6 = phi i64 [ 0, %for.cond1.preheader ], [ %indvar.next7, %for.inc25 ]
25  %arrayidx24 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @C, i64 0, i64 %indvar4, i64 %indvar6
26  store float 0.000000e+00, float* %arrayidx24, align 4
27  br label %for.body8
28
29for.body8:                                        ; preds = %for.body3, %for.body8
30  %indvar = phi i64 [ 0, %for.body3 ], [ %indvar.next, %for.body8 ]
31  %arrayidx16 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @A, i64 0, i64 %indvar4, i64 %indvar
32  %arrayidx20 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @B, i64 0, i64 %indvar, i64 %indvar6
33  %0 = load float, float* %arrayidx24, align 4
34  %1 = load float, float* %arrayidx16, align 4
35  %2 = load float, float* %arrayidx20, align 4
36  %mul = fmul float %1, %2
37  %add = fadd float %0, %mul
38  store float %add, float* %arrayidx24, align 4
39  %indvar.next = add i64 %indvar, 1
40  %exitcond = icmp ne i64 %indvar.next, 1536
41  br i1 %exitcond, label %for.body8, label %for.inc25
42
43for.inc25:                                        ; preds = %for.body8
44  %indvar.next7 = add i64 %indvar6, 1
45  %exitcond8 = icmp ne i64 %indvar.next7, 1536
46  br i1 %exitcond8, label %for.body3, label %for.inc28
47
48for.inc28:                                        ; preds = %for.inc25
49  %indvar.next5 = add i64 %indvar4, 1
50  %exitcond9 = icmp ne i64 %indvar.next5, 1536
51  br i1 %exitcond9, label %for.cond1.preheader, label %for.end30
52
53for.end30:                                        ; preds = %for.inc28
54  ret void
55}
56
57attributes #0 = { nounwind uwtable "less-precise-fpmad"="false" "frame-pointer"="all" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
58
59; CHECK: #pragma known-parallel
60; CHECK: for (int c0 = 0; c0 <= 47; c0 += 1)
61; CHECK:   for (int c1 = 0; c1 <= 47; c1 += 1)
62; CHECK:     for (int c2 = 0; c2 <= 31; c2 += 1)
63; CHECK:       for (int c3 = 0; c3 <= 7; c3 += 1)
64; CHECK:         // SIMD
65; CHECK:         for (int c4 = 0; c4 <= 3; c4 += 1)
66; CHECK:           Stmt_for_body3(32 * c0 + c2, 32 * c1 + 4 * c3 + c4);
67; CHECK: #pragma known-parallel
68; CHECK: for (int c0 = 0; c0 <= 47; c0 += 1)
69; CHECK:   for (int c1 = 0; c1 <= 47; c1 += 1)
70; CHECK:     for (int c2 = 0; c2 <= 47; c2 += 1)
71; CHECK:       for (int c3 = 0; c3 <= 31; c3 += 1)
72; CHECK:         for (int c4 = 0; c4 <= 7; c4 += 1)
73; CHECK:           for (int c5 = 0; c5 <= 31; c5 += 1)
74; CHECK:             // SIMD
75; CHECK:             for (int c6 = 0; c6 <= 3; c6 += 1)
76; CHECK:               Stmt_for_body8(32 * c0 + c3, 32 * c1 + 4 * c4 + c6, 32 * c2 + c5);
77
78; VEC16: {
79; VEC16:   #pragma known-parallel
80; VEC16:   for (int c0 = 0; c0 <= 47; c0 += 1)
81; VEC16:     for (int c1 = 0; c1 <= 47; c1 += 1)
82; VEC16:       for (int c2 = 0; c2 <= 31; c2 += 1)
83; VEC16:         for (int c3 = 0; c3 <= 1; c3 += 1)
84; VEC16:           // SIMD
85; VEC16:           for (int c4 = 0; c4 <= 15; c4 += 1)
86; VEC16:             Stmt_for_body3(32 * c0 + c2, 32 * c1 + 16 * c3 + c4);
87; VEC16:   #pragma known-parallel
88; VEC16:   for (int c0 = 0; c0 <= 47; c0 += 1)
89; VEC16:     for (int c1 = 0; c1 <= 47; c1 += 1)
90; VEC16:       for (int c2 = 0; c2 <= 47; c2 += 1)
91; VEC16:         for (int c3 = 0; c3 <= 31; c3 += 1)
92; VEC16:           for (int c4 = 0; c4 <= 1; c4 += 1)
93; VEC16:             for (int c5 = 0; c5 <= 31; c5 += 1)
94; VEC16:               // SIMD
95; VEC16:               for (int c6 = 0; c6 <= 15; c6 += 1)
96; VEC16:                 Stmt_for_body8(32 * c0 + c3, 32 * c1 + 16 * c4 + c6, 32 * c2 + c5);
97; VEC16: }
98
99
100!llvm.ident = !{!0}
101
102!0 = !{!"clang version 3.5.0 "}
103