110 lines · plain
1; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --filter "LV: Found an estimated cost of [0-9]+ for VF [0-9]+ For instruction:\s*store i8 %v6, ptr %out6"2; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+sse2 --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefix=SSE23; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefix=AVX14; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2 --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefix=AVX25; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512vl --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefix=AVX512DQ6; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512vl,+avx512bw --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefix=AVX512BW7; REQUIRES: asserts8 9target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"10target triple = "x86_64-unknown-linux-gnu"11 12@A = global [1024 x i8] zeroinitializer, align 12813@B = global [1024 x i8] zeroinitializer, align 12814 15define void @test() {16; SSE2-LABEL: 'test'17; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 118; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 119; SSE2: LV: Found an estimated cost of 57 for VF 2 For instruction: store i8 %v6, ptr %out6, align 120; SSE2: LV: Found an estimated cost of 112 for VF 4 For instruction: store i8 %v6, ptr %out6, align 121; SSE2: LV: Found an estimated cost of 225 for VF 8 For instruction: store i8 %v6, ptr %out6, align 122; SSE2: LV: Found an estimated cost of 456 for VF 16 For instruction: store i8 %v6, ptr %out6, align 123;24; AVX1-LABEL: 'test'25; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 126; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 127; AVX1: LV: Found an estimated cost of 34 for VF 2 For instruction: store i8 %v6, ptr %out6, align 128; AVX1: LV: Found an estimated cost of 63 for VF 4 For instruction: store i8 %v6, ptr %out6, align 129; AVX1: LV: Found an estimated cost of 119 for VF 8 For instruction: store i8 %v6, ptr %out6, align 130; AVX1: LV: Found an estimated cost of 232 for VF 16 For instruction: store i8 %v6, ptr %out6, align 131; AVX1: LV: Found an estimated cost of 469 for VF 32 For instruction: store i8 %v6, ptr %out6, align 132;33; AVX2-LABEL: 'test'34; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 135; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 136; AVX2: LV: Found an estimated cost of 34 for VF 2 For instruction: store i8 %v6, ptr %out6, align 137; AVX2: LV: Found an estimated cost of 63 for VF 4 For instruction: store i8 %v6, ptr %out6, align 138; AVX2: LV: Found an estimated cost of 119 for VF 8 For instruction: store i8 %v6, ptr %out6, align 139; AVX2: LV: Found an estimated cost of 232 for VF 16 For instruction: store i8 %v6, ptr %out6, align 140; AVX2: LV: Found an estimated cost of 469 for VF 32 For instruction: store i8 %v6, ptr %out6, align 141;42; AVX512DQ-LABEL: 'test'43; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 144; AVX512DQ: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 145; AVX512DQ: LV: Found an estimated cost of 34 for VF 2 For instruction: store i8 %v6, ptr %out6, align 146; AVX512DQ: LV: Found an estimated cost of 63 for VF 4 For instruction: store i8 %v6, ptr %out6, align 147; AVX512DQ: LV: Found an estimated cost of 121 for VF 8 For instruction: store i8 %v6, ptr %out6, align 148; AVX512DQ: LV: Found an estimated cost of 234 for VF 16 For instruction: store i8 %v6, ptr %out6, align 149; AVX512DQ: LV: Found an estimated cost of 470 for VF 32 For instruction: store i8 %v6, ptr %out6, align 150; AVX512DQ: LV: Found an estimated cost of 945 for VF 64 For instruction: store i8 %v6, ptr %out6, align 151;52; AVX512BW-LABEL: 'test'53; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 154; AVX512BW: LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v6, ptr %out6, align 155; AVX512BW: LV: Found an estimated cost of 22 for VF 2 For instruction: store i8 %v6, ptr %out6, align 156; AVX512BW: LV: Found an estimated cost of 46 for VF 4 For instruction: store i8 %v6, ptr %out6, align 157; AVX512BW: LV: Found an estimated cost of 118 for VF 8 For instruction: store i8 %v6, ptr %out6, align 158; AVX512BW: LV: Found an estimated cost of 236 for VF 16 For instruction: store i8 %v6, ptr %out6, align 159; AVX512BW: LV: Found an estimated cost of 472 for VF 32 For instruction: store i8 %v6, ptr %out6, align 160; AVX512BW: LV: Found an estimated cost of 826 for VF 64 For instruction: store i8 %v6, ptr %out6, align 161;62entry:63 br label %for.body64 65for.body:66 %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]67 68 %iv.0 = add nuw nsw i64 %iv, 069 %iv.1 = add nuw nsw i64 %iv, 170 %iv.2 = add nuw nsw i64 %iv, 271 %iv.3 = add nuw nsw i64 %iv, 372 %iv.4 = add nuw nsw i64 %iv, 473 %iv.5 = add nuw nsw i64 %iv, 574 %iv.6 = add nuw nsw i64 %iv, 675 76 %in = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.077 %v = load i8, ptr %in78 79 %v0 = add i8 %v, 080 %v1 = add i8 %v, 181 %v2 = add i8 %v, 282 %v3 = add i8 %v, 383 %v4 = add i8 %v, 484 %v5 = add i8 %v, 585 %v6 = add i8 %v, 686 87 %out0 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.088 %out1 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.189 %out2 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.290 %out3 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.391 %out4 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.492 %out5 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.593 %out6 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.694 95 store i8 %v0, ptr %out096 store i8 %v1, ptr %out197 store i8 %v2, ptr %out298 store i8 %v3, ptr %out399 store i8 %v4, ptr %out4100 store i8 %v5, ptr %out5101 store i8 %v6, ptr %out6102 103 %iv.next = add nuw nsw i64 %iv.0, 7104 %cmp = icmp ult i64 %iv.next, 1024105 br i1 %cmp, label %for.body, label %for.cond.cleanup106 107for.cond.cleanup:108 ret void109}110