1; REQUIRES: asserts
2
3; RUN: opt -loop-vectorize -debug-only=loop-vectorize -force-vector-interleave=1 -force-vector-width=4 -vplan-print-in-dot-format -disable-output %s 2>&1 | FileCheck %s
4
5target 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"
6
7; Verify that -vplan-print-in-dot-format option works.
8
9define void @print_call_and_memory(i64 %n, float* noalias %y, float* noalias %x) nounwind uwtable {
10; CHECK:      digraph VPlan {
11; CHECK-NEXT:  graph [labelloc=t, fontsize=30; label="Vectorization Plan\nInitial VPlan for VF=\{4\},UF\>=1"]
12; CHECK-NEXT:  node [shape=rect, fontname=Courier, fontsize=30]
13; CHECK-NEXT:  edge [fontname=Courier, fontsize=30]
14; CHECK-NEXT:  compound=true
15; CHECK-NEXT:  N0 [label =
16; CHECK-NEXT:    "vector.ph:\l" +
17; CHECK-NEXT:    "Successor(s): vector loop\l"
18; CHECK-NEXT:  ]
19; CHECK-NEXT:  N0 -> N1 [ label="" lhead=cluster_N2]
20; CHECK-NEXT:  subgraph cluster_N2 {
21; CHECK-NEXT:    fontname=Courier
22; CHECK-NEXT:    label="\<x1\> vector loop"
23; CHECK-NEXT:    N1 [label =
24; CHECK-NEXT:    "vector.body:\l" +
25; CHECK-NEXT:    "  EMIT vp\<[[CAN_IV:%.+]]\> = CANONICAL-INDUCTION\l" +
26; CHECK-NEXT:    "  vp\<[[STEPS:%.+]]\> = SCALAR-STEPS vp\<[[CAN_IV]]\>, ir\<0\>, ir\<1\>\l" +
27; CHECK-NEXT:    "  CLONE ir\<%arrayidx\> = getelementptr ir\<%y\>, vp\<[[STEPS]]\>\l" +
28; CHECK-NEXT:    "  WIDEN ir\<%lv\> = load ir\<%arrayidx\>\l" +
29; CHECK-NEXT:    "  WIDEN-CALL ir\<%call\> = call @llvm.sqrt.f32(ir\<%lv\>)\l" +
30; CHECK-NEXT:    "  CLONE ir\<%arrayidx2\> = getelementptr ir\<%x\>, vp\<[[STEPS]]\>\l" +
31; CHECK-NEXT:    "  WIDEN store ir\<%arrayidx2\>, ir\<%call\>\l" +
32; CHECK-NEXT:    "  EMIT vp\<[[CAN_IV_NEXT:%.+]]\> = VF * UF +(nuw) vp\<[[CAN_IV]]\>\l" +
33; CHECK-NEXT:    "  EMIT branch-on-count vp\<[[CAN_IV_NEXT]]\> vp\<{{.+}}\>\l" +
34; CHECK-NEXT:    "No successors\l"
35; CHECK-NEXT:  ]
36;
37entry:
38  %cmp6 = icmp sgt i64 %n, 0
39  br i1 %cmp6, label %for.body, label %for.end
40
41for.body:                                         ; preds = %entry, %for.body
42  %iv = phi i64 [ %iv.next, %for.body ], [ 0, %entry ]
43  %arrayidx = getelementptr inbounds float, float* %y, i64 %iv
44  %lv = load float, float* %arrayidx, align 4
45  %call = tail call float @llvm.sqrt.f32(float %lv) nounwind readnone
46  %arrayidx2 = getelementptr inbounds float, float* %x, i64 %iv
47  store float %call, float* %arrayidx2, align 4
48  %iv.next = add i64 %iv, 1
49  %exitcond = icmp eq i64 %iv.next, %n
50  br i1 %exitcond, label %for.end, label %for.body
51
52for.end:                                          ; preds = %for.body, %entry
53  ret void
54}
55
56declare float @llvm.sqrt.f32(float) nounwind readnone
57