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=+sse42 --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,SSE42
3 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX1
4 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2,-fast-gather --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX2-SLOWGATHER
5 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2,+fast-gather --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX2-FASTGATHER
6 ; RUN: opt -loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512bw --debug-only=loop-vectorize < %s 2>&1 | FileCheck %s --check-prefixes=CHECK,AVX512
10 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
11 target triple = "x86_64-unknown-linux-gnu"
13 @A = global [1024 x i8] zeroinitializer, align 128
14 @B = global [1024 x i64] zeroinitializer, align 128
16 ; CHECK: LV: Checking a loop in "test"
18 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
19 ; SSE2: LV: Found an estimated cost of 2 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
20 ; SSE2: LV: Found an estimated cost of 5 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
21 ; SSE2: LV: Found an estimated cost of 10 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
22 ; SSE2: LV: Found an estimated cost of 20 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
24 ; SSE42: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
25 ; SSE42: LV: Found an estimated cost of 2 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
26 ; SSE42: LV: Found an estimated cost of 5 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
27 ; SSE42: LV: Found an estimated cost of 10 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
28 ; SSE42: LV: Found an estimated cost of 20 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
30 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
31 ; AVX1: LV: Found an estimated cost of 8 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
32 ; AVX1: LV: Found an estimated cost of 8 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
33 ; AVX1: LV: Found an estimated cost of 16 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
34 ; AVX1: LV: Found an estimated cost of 32 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
35 ; AVX1: LV: Found an estimated cost of 64 for VF 32 For instruction: store i64 %valB, i64* %out, align 8
37 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
38 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 8 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
39 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 8 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
40 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 16 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
41 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 32 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
42 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 64 for VF 32 For instruction: store i64 %valB, i64* %out, align 8
44 ; AVX2-FASTGATHER: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
45 ; AVX2-FASTGATHER: LV: Found an estimated cost of 8 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
46 ; AVX2-FASTGATHER: LV: Found an estimated cost of 8 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
47 ; AVX2-FASTGATHER: LV: Found an estimated cost of 16 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
48 ; AVX2-FASTGATHER: LV: Found an estimated cost of 32 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
49 ; AVX2-FASTGATHER: LV: Found an estimated cost of 64 for VF 32 For instruction: store i64 %valB, i64* %out, align 8
51 ; AVX512: LV: Found an estimated cost of 1 for VF 1 For instruction: store i64 %valB, i64* %out, align 8
52 ; AVX512: LV: Found an estimated cost of 1 for VF 2 For instruction: store i64 %valB, i64* %out, align 8
53 ; AVX512: LV: Found an estimated cost of 1 for VF 4 For instruction: store i64 %valB, i64* %out, align 8
54 ; AVX512: LV: Found an estimated cost of 1 for VF 8 For instruction: store i64 %valB, i64* %out, align 8
55 ; AVX512: LV: Found an estimated cost of 2 for VF 16 For instruction: store i64 %valB, i64* %out, align 8
56 ; AVX512: LV: Found an estimated cost of 4 for VF 32 For instruction: store i64 %valB, i64* %out, align 8
57 ; AVX512: LV: Found an estimated cost of 8 for VF 64 For instruction: store i64 %valB, i64* %out, align 8
59 ; CHECK-NOT: LV: Found an estimated cost of {{[0-9]+}} for VF {{[0-9]+}} For instruction: store i64 %valB, i64* %out, align 8
60 define void @test([1024 x i64]* %C) {
65 %iv = phi i64 [ 0, %entry ], [ %iv.next, %end ]
67 %inB = getelementptr inbounds [1024 x i64], [1024 x i64]* @B, i64 0, i64 %iv
68 %valB = load i64, i64* %inB
70 %inA = getelementptr inbounds [1024 x i8], [1024 x i8]* @A, i64 0, i64 %iv
71 %valA = load i8, i8* %inA
72 %canStore = icmp ne i8 %valA, 0
73 br i1 %canStore, label %store, label %mask
76 %out = getelementptr inbounds [1024 x i64], [1024 x i64]* %C, i64 0, i64 %iv
77 store i64 %valB, i64* %out
84 %iv.next = add nuw nsw i64 %iv, 1
85 %cmp = icmp ult i64 %iv.next, 1024
86 br i1 %cmp, label %for.body, label %for.cond.cleanup