1; RUN: opt -S %loadPolly -basicaa -polly-opt-isl -polly-vectorizer=polly -polly-ast -analyze < %s | FileCheck %s
2; RUN: opt -S %loadPolly -basicaa -polly-opt-isl -polly-vectorizer=stripmine -polly-ast -analyze < %s | FileCheck %s
3
4; RUN: opt -S %loadPolly -basicaa -polly-opt-isl \
5; RUN:                   -polly-vectorizer=polly -polly-ast -analyze \
6; RUN:                   -polly-prevect-width=16 < %s | \
7; RUN:                   FileCheck %s -check-prefix=VEC16
8
9target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
10
11@C = common global [1536 x [1536 x float]] zeroinitializer, align 16
12@A = common global [1536 x [1536 x float]] zeroinitializer, align 16
13@B = common global [1536 x [1536 x float]] zeroinitializer, align 16
14
15; Function Attrs: nounwind uwtable
16define void @foo() #0 {
17entry:
18  br label %entry.split
19
20entry.split:                                      ; preds = %entry
21  br label %for.cond1.preheader
22
23for.cond1.preheader:                              ; preds = %entry.split, %for.inc28
24  %indvar4 = phi i64 [ 0, %entry.split ], [ %indvar.next5, %for.inc28 ]
25  br label %for.body3
26
27for.body3:                                        ; preds = %for.cond1.preheader, %for.inc25
28  %indvar6 = phi i64 [ 0, %for.cond1.preheader ], [ %indvar.next7, %for.inc25 ]
29  %arrayidx24 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @C, i64 0, i64 %indvar4, i64 %indvar6
30  store float 0.000000e+00, float* %arrayidx24, align 4
31  br label %for.body8
32
33for.body8:                                        ; preds = %for.body3, %for.body8
34  %indvar = phi i64 [ 0, %for.body3 ], [ %indvar.next, %for.body8 ]
35  %arrayidx16 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @A, i64 0, i64 %indvar4, i64 %indvar
36  %arrayidx20 = getelementptr [1536 x [1536 x float]], [1536 x [1536 x float]]* @B, i64 0, i64 %indvar, i64 %indvar6
37  %0 = load float, float* %arrayidx24, align 4
38  %1 = load float, float* %arrayidx16, align 4
39  %2 = load float, float* %arrayidx20, align 4
40  %mul = fmul float %1, %2
41  %add = fadd float %0, %mul
42  store float %add, float* %arrayidx24, align 4
43  %indvar.next = add i64 %indvar, 1
44  %exitcond = icmp ne i64 %indvar.next, 1536
45  br i1 %exitcond, label %for.body8, label %for.inc25
46
47for.inc25:                                        ; preds = %for.body8
48  %indvar.next7 = add i64 %indvar6, 1
49  %exitcond8 = icmp ne i64 %indvar.next7, 1536
50  br i1 %exitcond8, label %for.body3, label %for.inc28
51
52for.inc28:                                        ; preds = %for.inc25
53  %indvar.next5 = add i64 %indvar4, 1
54  %exitcond9 = icmp ne i64 %indvar.next5, 1536
55  br i1 %exitcond9, label %for.cond1.preheader, label %for.end30
56
57for.end30:                                        ; preds = %for.inc28
58  ret void
59}
60
61attributes #0 = { nounwind uwtable "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" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
62
63; CHECK: #pragma known-parallel
64; CHECK: for (int c0 = 0; c0 <= 47; c0 += 1)
65; CHECK:   for (int c1 = 0; c1 <= 47; c1 += 1)
66; CHECK:     for (int c2 = 0; c2 <= 31; c2 += 1)
67; CHECK:       for (int c3 = 0; c3 <= 7; c3 += 1)
68; CHECK:         // SIMD
69; CHECK:         for (int c4 = 0; c4 <= 3; c4 += 1)
70; CHECK:           Stmt_for_body3(32 * c0 + c2, 32 * c1 + 4 * c3 + c4);
71; CHECK: #pragma known-parallel
72; CHECK: for (int c0 = 0; c0 <= 47; c0 += 1)
73; CHECK:   for (int c1 = 0; c1 <= 47; c1 += 1)
74; CHECK:     for (int c2 = 0; c2 <= 47; c2 += 1)
75; CHECK:       for (int c3 = 0; c3 <= 31; c3 += 1)
76; CHECK:         for (int c4 = 0; c4 <= 7; c4 += 1)
77; CHECK:           for (int c5 = 0; c5 <= 31; c5 += 1)
78; CHECK:             // SIMD
79; CHECK:             for (int c6 = 0; c6 <= 3; c6 += 1)
80; CHECK:               Stmt_for_body8(32 * c0 + c3, 32 * c1 + 4 * c4 + c6, 32 * c2 + c5);
81
82; VEC16: {
83; VEC16:   #pragma known-parallel
84; VEC16:   for (int c0 = 0; c0 <= 47; c0 += 1)
85; VEC16:     for (int c1 = 0; c1 <= 47; c1 += 1)
86; VEC16:       for (int c2 = 0; c2 <= 31; c2 += 1)
87; VEC16:         for (int c3 = 0; c3 <= 1; c3 += 1)
88; VEC16:           // SIMD
89; VEC16:           for (int c4 = 0; c4 <= 15; c4 += 1)
90; VEC16:             Stmt_for_body3(32 * c0 + c2, 32 * c1 + 16 * c3 + c4);
91; VEC16:   #pragma known-parallel
92; VEC16:   for (int c0 = 0; c0 <= 47; c0 += 1)
93; VEC16:     for (int c1 = 0; c1 <= 47; c1 += 1)
94; VEC16:       for (int c2 = 0; c2 <= 47; c2 += 1)
95; VEC16:         for (int c3 = 0; c3 <= 31; c3 += 1)
96; VEC16:           for (int c4 = 0; c4 <= 1; c4 += 1)
97; VEC16:             for (int c5 = 0; c5 <= 31; c5 += 1)
98; VEC16:               // SIMD
99; VEC16:               for (int c6 = 0; c6 <= 15; c6 += 1)
100; VEC16:                 Stmt_for_body8(32 * c0 + c3, 32 * c1 + 16 * c4 + c6, 32 * c2 + c5);
101; VEC16: }
102
103
104!llvm.ident = !{!0}
105
106!0 = !{!"clang version 3.5.0 "}
107