89 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 i32 %valB, ptr %out, align 4"2; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+sse2 --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=SSE23; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+sse4.2 --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=SSE424; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=AVX15; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2,-fast-gather --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=AVX26; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx2,+fast-gather --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=AVX27; RUN: opt -passes=loop-vectorize -vectorizer-maximize-bandwidth -S -mattr=+avx512bw --debug-only=loop-vectorize --disable-output < %s 2>&1 | FileCheck %s --check-prefixes=AVX5128 9; REQUIRES: asserts10 11target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"12target triple = "x86_64-unknown-linux-gnu"13 14@A = global [1024 x i8] zeroinitializer, align 12815@B = global [1024 x i32] zeroinitializer, align 12816@C = global [1024 x i32] zeroinitializer, align 12817 18define void @test() {19; SSE2-LABEL: 'test'20; SSE2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i32 %valB, ptr %out, align 421; SSE2: LV: Found an estimated cost of 2 for VF 2 For instruction: store i32 %valB, ptr %out, align 422; SSE2: LV: Found an estimated cost of 5 for VF 4 For instruction: store i32 %valB, ptr %out, align 423; SSE2: LV: Found an estimated cost of 11 for VF 8 For instruction: store i32 %valB, ptr %out, align 424; SSE2: LV: Found an estimated cost of 22 for VF 16 For instruction: store i32 %valB, ptr %out, align 425;26; SSE42-LABEL: 'test'27; SSE42: LV: Found an estimated cost of 1 for VF 1 For instruction: store i32 %valB, ptr %out, align 428; SSE42: LV: Found an estimated cost of 2 for VF 2 For instruction: store i32 %valB, ptr %out, align 429; SSE42: LV: Found an estimated cost of 4 for VF 4 For instruction: store i32 %valB, ptr %out, align 430; SSE42: LV: Found an estimated cost of 8 for VF 8 For instruction: store i32 %valB, ptr %out, align 431; SSE42: LV: Found an estimated cost of 16 for VF 16 For instruction: store i32 %valB, ptr %out, align 432;33; AVX1-LABEL: 'test'34; AVX1: LV: Found an estimated cost of 1 for VF 1 For instruction: store i32 %valB, ptr %out, align 435; AVX1: LV: Found an estimated cost of 2 for VF 2 For instruction: store i32 %valB, ptr %out, align 436; AVX1: LV: Found an estimated cost of 4 for VF 4 For instruction: store i32 %valB, ptr %out, align 437; AVX1: LV: Found an estimated cost of 8 for VF 8 For instruction: store i32 %valB, ptr %out, align 438; AVX1: LV: Found an estimated cost of 17 for VF 16 For instruction: store i32 %valB, ptr %out, align 439; AVX1: LV: Found an estimated cost of 34 for VF 32 For instruction: store i32 %valB, ptr %out, align 440;41; AVX2-LABEL: 'test'42; AVX2: LV: Found an estimated cost of 1 for VF 1 For instruction: store i32 %valB, ptr %out, align 443; AVX2: LV: Found an estimated cost of 2 for VF 2 For instruction: store i32 %valB, ptr %out, align 444; AVX2: LV: Found an estimated cost of 4 for VF 4 For instruction: store i32 %valB, ptr %out, align 445; AVX2: LV: Found an estimated cost of 8 for VF 8 For instruction: store i32 %valB, ptr %out, align 446; AVX2: LV: Found an estimated cost of 17 for VF 16 For instruction: store i32 %valB, ptr %out, align 447; AVX2: LV: Found an estimated cost of 34 for VF 32 For instruction: store i32 %valB, ptr %out, align 448;49; AVX512-LABEL: 'test'50; AVX512: LV: Found an estimated cost of 1 for VF 1 For instruction: store i32 %valB, ptr %out, align 451; AVX512: LV: Found an estimated cost of 5 for VF 2 For instruction: store i32 %valB, ptr %out, align 452; AVX512: LV: Found an estimated cost of 10 for VF 4 For instruction: store i32 %valB, ptr %out, align 453; AVX512: LV: Found an estimated cost of 10 for VF 8 For instruction: store i32 %valB, ptr %out, align 454; AVX512: LV: Found an estimated cost of 18 for VF 16 For instruction: store i32 %valB, ptr %out, align 455; AVX512: LV: Found an estimated cost of 36 for VF 32 For instruction: store i32 %valB, ptr %out, align 456; AVX512: LV: Found an estimated cost of 72 for VF 64 For instruction: store i32 %valB, ptr %out, align 457;58entry:59 br label %for.body60 61for.body:62 %iv = phi i64 [ 0, %entry ], [ %iv.next, %end ]63 64 %inB = getelementptr inbounds [1024 x i32], ptr @B, i64 0, i64 %iv65 %valB = load i32, ptr %inB66 67 %inA = getelementptr inbounds [1024 x i8], ptr @A, i64 0, i64 %iv68 %valA = load i8, ptr %inA69 %canStore = icmp ne i8 %valA, 070 br i1 %canStore, label %store, label %mask71 72store:73 %valA.ext = sext i8 %valA to i6474 %out = getelementptr inbounds [1024 x i32], ptr @C, i64 0, i64 %valA.ext75 store i32 %valB, ptr %out76 br label %end77 78mask:79 br label %end80 81end:82 %iv.next = add nuw nsw i64 %iv, 183 %cmp = icmp ult i64 %iv.next, 102484 br i1 %cmp, label %for.body, label %for.cond.cleanup85 86for.cond.cleanup:87 ret void88}89