brintos

brintos / llvm-project-archived public Read only

0
0
Text · 6.2 KiB · 1c07920 Raw
102 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 %v4, ptr %out4"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 %v4, ptr %out4, align 118; SSE2:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 119; SSE2:  LV: Found an estimated cost of 44 for VF 2 For instruction: store i8 %v4, ptr %out4, align 120; SSE2:  LV: Found an estimated cost of 87 for VF 4 For instruction: store i8 %v4, ptr %out4, align 121; SSE2:  LV: Found an estimated cost of 178 for VF 8 For instruction: store i8 %v4, ptr %out4, align 122; SSE2:  LV: Found an estimated cost of 360 for VF 16 For instruction: store i8 %v4, ptr %out4, align 123;24; AVX1-LABEL: 'test'25; AVX1:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 126; AVX1:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 127; AVX1:  LV: Found an estimated cost of 24 for VF 2 For instruction: store i8 %v4, ptr %out4, align 128; AVX1:  LV: Found an estimated cost of 46 for VF 4 For instruction: store i8 %v4, ptr %out4, align 129; AVX1:  LV: Found an estimated cost of 84 for VF 8 For instruction: store i8 %v4, ptr %out4, align 130; AVX1:  LV: Found an estimated cost of 166 for VF 16 For instruction: store i8 %v4, ptr %out4, align 131; AVX1:  LV: Found an estimated cost of 335 for VF 32 For instruction: store i8 %v4, ptr %out4, align 132;33; AVX2-LABEL: 'test'34; AVX2:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 135; AVX2:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 136; AVX2:  LV: Found an estimated cost of 24 for VF 2 For instruction: store i8 %v4, ptr %out4, align 137; AVX2:  LV: Found an estimated cost of 46 for VF 4 For instruction: store i8 %v4, ptr %out4, align 138; AVX2:  LV: Found an estimated cost of 84 for VF 8 For instruction: store i8 %v4, ptr %out4, align 139; AVX2:  LV: Found an estimated cost of 166 for VF 16 For instruction: store i8 %v4, ptr %out4, align 140; AVX2:  LV: Found an estimated cost of 335 for VF 32 For instruction: store i8 %v4, ptr %out4, align 141;42; AVX512DQ-LABEL: 'test'43; AVX512DQ:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 144; AVX512DQ:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 145; AVX512DQ:  LV: Found an estimated cost of 24 for VF 2 For instruction: store i8 %v4, ptr %out4, align 146; AVX512DQ:  LV: Found an estimated cost of 46 for VF 4 For instruction: store i8 %v4, ptr %out4, align 147; AVX512DQ:  LV: Found an estimated cost of 87 for VF 8 For instruction: store i8 %v4, ptr %out4, align 148; AVX512DQ:  LV: Found an estimated cost of 166 for VF 16 For instruction: store i8 %v4, ptr %out4, align 149; AVX512DQ:  LV: Found an estimated cost of 336 for VF 32 For instruction: store i8 %v4, ptr %out4, align 150; AVX512DQ:  LV: Found an estimated cost of 675 for VF 64 For instruction: store i8 %v4, ptr %out4, align 151;52; AVX512BW-LABEL: 'test'53; AVX512BW:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 154; AVX512BW:  LV: Found an estimated cost of 1 for VF 1 For instruction: store i8 %v4, ptr %out4, align 155; AVX512BW:  LV: Found an estimated cost of 15 for VF 2 For instruction: store i8 %v4, ptr %out4, align 156; AVX512BW:  LV: Found an estimated cost of 31 for VF 4 For instruction: store i8 %v4, ptr %out4, align 157; AVX512BW:  LV: Found an estimated cost of 79 for VF 8 For instruction: store i8 %v4, ptr %out4, align 158; AVX512BW:  LV: Found an estimated cost of 158 for VF 16 For instruction: store i8 %v4, ptr %out4, align 159; AVX512BW:  LV: Found an estimated cost of 237 for VF 32 For instruction: store i8 %v4, ptr %out4, align 160; AVX512BW:  LV: Found an estimated cost of 395 for VF 64 For instruction: store i8 %v4, ptr %out4, 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 74  %in = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv.075  %v = load i8, ptr %in76 77  %v0 = add i8 %v, 078  %v1 = add i8 %v, 179  %v2 = add i8 %v, 280  %v3 = add i8 %v, 381  %v4 = add i8 %v, 482 83  %out0 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.084  %out1 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.185  %out2 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.286  %out3 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.387  %out4 = getelementptr inbounds [1024 x i8], ptr @B, i64 0, i64 %iv.488 89  store i8 %v0, ptr %out090  store i8 %v1, ptr %out191  store i8 %v2, ptr %out292  store i8 %v3, ptr %out393  store i8 %v4, ptr %out494 95  %iv.next = add nuw nsw i64 %iv.0, 596  %cmp = icmp ult i64 %iv.next, 102497  br i1 %cmp, label %for.body, label %for.cond.cleanup98 99for.cond.cleanup:100  ret void101}102