1 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+sse2 --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,SSE2
2 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX1
3 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2 --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX2
4 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512bw,+avx512vl --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX512
7 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
8 target triple = "x86_64-unknown-linux-gnu"
10 @A = global [1024 x i32] zeroinitializer, align 128
11 @B = global [1024 x i8] zeroinitializer, align 128
13 ; CHECK: LV: Checking a loop in "test"
15 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i32, i32* %in0, align 4
16 ; SSE2: LV: Found an estimated cost of 8 for VF 2 For instruction: %v0 = load i32, i32* %in0, align 4
17 ; SSE2: LV: Found an estimated cost of 17 for VF 4 For instruction: %v0 = load i32, i32* %in0, align 4
18 ; SSE2: LV: Found an estimated cost of 34 for VF 8 For instruction: %v0 = load i32, i32* %in0, align 4
19 ; SSE2: LV: Found an estimated cost of 68 for VF 16 For instruction: %v0 = load i32, i32* %in0, align 4
21 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i32, i32* %in0, align 4
22 ; AVX1: LV: Found an estimated cost of 7 for VF 2 For instruction: %v0 = load i32, i32* %in0, align 4
23 ; AVX1: LV: Found an estimated cost of 11 for VF 4 For instruction: %v0 = load i32, i32* %in0, align 4
24 ; AVX1: LV: Found an estimated cost of 25 for VF 8 For instruction: %v0 = load i32, i32* %in0, align 4
25 ; AVX1: LV: Found an estimated cost of 50 for VF 16 For instruction: %v0 = load i32, i32* %in0, align 4
26 ; AVX1: LV: Found an estimated cost of 100 for VF 32 For instruction: %v0 = load i32, i32* %in0, align 4
28 ; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i32, i32* %in0, align 4
29 ; AVX2: LV: Found an estimated cost of 4 for VF 2 For instruction: %v0 = load i32, i32* %in0, align 4
30 ; AVX2: LV: Found an estimated cost of 3 for VF 4 For instruction: %v0 = load i32, i32* %in0, align 4
31 ; AVX2: LV: Found an estimated cost of 6 for VF 8 For instruction: %v0 = load i32, i32* %in0, align 4
32 ; AVX2: LV: Found an estimated cost of 11 for VF 16 For instruction: %v0 = load i32, i32* %in0, align 4
33 ; AVX2: LV: Found an estimated cost of 23 for VF 32 For instruction: %v0 = load i32, i32* %in0, align 4
35 ; AVX512: LV: Found an estimated cost of 1 for VF 1 For instruction: %v0 = load i32, i32* %in0, align 4
36 ; AVX512: LV: Found an estimated cost of 1 for VF 2 For instruction: %v0 = load i32, i32* %in0, align 4
37 ; AVX512: LV: Found an estimated cost of 1 for VF 4 For instruction: %v0 = load i32, i32* %in0, align 4
38 ; AVX512: LV: Found an estimated cost of 2 for VF 8 For instruction: %v0 = load i32, i32* %in0, align 4
39 ; AVX512: LV: Found an estimated cost of 3 for VF 16 For instruction: %v0 = load i32, i32* %in0, align 4
40 ; AVX512: LV: Found an estimated cost of 21 for VF 32 For instruction: %v0 = load i32, i32* %in0, align 4
41 ; AVX512: LV: Found an estimated cost of 78 for VF 64 For instruction: %v0 = load i32, i32* %in0, align 4
43 ; CHECK-NOT: LV: Found an estimated cost of {{[0-9]+}} for VF {{[0-9]+}} For instruction: %v0 = load i32, i32* %in0, align 4
50 %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
52 %iv.0 = add nuw nsw i64 %iv, 0
54 %in0 = getelementptr inbounds [1024 x i32], [1024 x i32]* @A, i64 0, i64 %iv.0
56 %v0 = load i32, i32* %in0
58 %reduce.add.0 = add i32 %v0, 0
59 %reduce.add.1 = add i32 %reduce.add.0, 0
61 %reduce.add.1.narrow = trunc i32 %reduce.add.1 to i8
63 %out = getelementptr inbounds [1024 x i8], [1024 x i8]* @B, i64 0, i64 %iv.0
64 store i8 %reduce.add.1.narrow, i8* %out
66 %iv.next = add nuw nsw i64 %iv, 3
67 %cmp = icmp ult i64 %iv.next, 1024
68 br i1 %cmp, label %for.body, label %for.cond.cleanup