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 i16] zeroinitializer, align 128
15 @C = global [1024 x i16] zeroinitializer, align 128
17 ; CHECK: LV: Checking a loop in "test"
19 ; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
20 ; SSE2: LV: Found an estimated cost of 24 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
21 ; SSE2: LV: Found an estimated cost of 48 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
22 ; SSE2: LV: Found an estimated cost of 96 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
23 ; SSE2: LV: Found an estimated cost of 192 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
25 ; SSE42: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
26 ; SSE42: LV: Found an estimated cost of 24 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
27 ; SSE42: LV: Found an estimated cost of 48 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
28 ; SSE42: LV: Found an estimated cost of 96 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
29 ; SSE42: LV: Found an estimated cost of 192 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
31 ; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
32 ; AVX1: LV: Found an estimated cost of 24 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
33 ; AVX1: LV: Found an estimated cost of 48 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
34 ; AVX1: LV: Found an estimated cost of 96 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
35 ; AVX1: LV: Found an estimated cost of 194 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
36 ; AVX1: LV: Found an estimated cost of 388 for VF 32 For instruction: %valB = load i16, i16* %inB, align 2
38 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
39 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 4 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
40 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 8 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
41 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 16 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
42 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 34 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
43 ; AVX2-SLOWGATHER: LV: Found an estimated cost of 68 for VF 32 For instruction: %valB = load i16, i16* %inB, align 2
45 ; AVX2-FASTGATHER: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
46 ; AVX2-FASTGATHER: LV: Found an estimated cost of 6 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
47 ; AVX2-FASTGATHER: LV: Found an estimated cost of 14 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
48 ; AVX2-FASTGATHER: LV: Found an estimated cost of 28 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
49 ; AVX2-FASTGATHER: LV: Found an estimated cost of 58 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
50 ; AVX2-FASTGATHER: LV: Found an estimated cost of 116 for VF 32 For instruction: %valB = load i16, i16* %inB, align 2
52 ; AVX512: LV: Found an estimated cost of 1 for VF 1 For instruction: %valB = load i16, i16* %inB, align 2
53 ; AVX512: LV: Found an estimated cost of 6 for VF 2 For instruction: %valB = load i16, i16* %inB, align 2
54 ; AVX512: LV: Found an estimated cost of 14 for VF 4 For instruction: %valB = load i16, i16* %inB, align 2
55 ; AVX512: LV: Found an estimated cost of 30 for VF 8 For instruction: %valB = load i16, i16* %inB, align 2
56 ; AVX512: LV: Found an estimated cost of 62 for VF 16 For instruction: %valB = load i16, i16* %inB, align 2
57 ; AVX512: LV: Found an estimated cost of 124 for VF 32 For instruction: %valB = load i16, i16* %inB, align 2
58 ; AVX512: LV: Found an estimated cost of 248 for VF 64 For instruction: %valB = load i16, i16* %inB, align 2
60 ; CHECK-NOT: LV: Found an estimated cost of {{[0-9]+}} for VF {{[0-9]+}} For instruction: %valB = load i16, i16* %inB, align 2
66 %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
68 %inA = getelementptr inbounds [1024 x i8], [1024 x i8]* @A, i64 0, i64 %iv
69 %valA = load i8, i8* %inA
70 %valA.ext = sext i8 %valA to i64
72 %inB = getelementptr inbounds [1024 x i16], [1024 x i16]* @B, i64 0, i64 %valA.ext
73 %valB = load i16, i16* %inB
75 %out = getelementptr inbounds [1024 x i16], [1024 x i16]* @C, i64 0, i64 %iv
76 store i16 %valB, i16* %out
78 %iv.next = add nuw nsw i64 %iv, 1
79 %cmp = icmp ult i64 %iv.next, 1024
80 br i1 %cmp, label %for.body, label %for.cond.cleanup