46 lines · plain
1; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py2; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -cost-kind=all -mtriple=aarch64 < %s | FileCheck %s3 4target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"5 6define void @reduce() {7; CHECK-LABEL: 'reduce'8; CHECK-NEXT: Cost Model: Found costs of 2 for: %V1i8 = call i8 @llvm.vector.reduce.add.v1i8(<1 x i8> undef)9; CHECK-NEXT: Cost Model: Found costs of 2 for: %V2i8 = call i8 @llvm.vector.reduce.add.v2i8(<2 x i8> undef)10; CHECK-NEXT: Cost Model: Found costs of 2 for: %V3i8 = call i8 @llvm.vector.reduce.add.v3i8(<3 x i8> undef)11; CHECK-NEXT: Cost Model: Found costs of 2 for: %V4i8 = call i8 @llvm.vector.reduce.add.v4i8(<4 x i8> undef)12; CHECK-NEXT: Cost Model: Found costs of 2 for: %V8i8 = call i8 @llvm.vector.reduce.add.v8i8(<8 x i8> undef)13; CHECK-NEXT: Cost Model: Found costs of 2 for: %V16i8 = call i8 @llvm.vector.reduce.add.v16i8(<16 x i8> undef)14; CHECK-NEXT: Cost Model: Found costs of 3 for: %V32i8 = call i8 @llvm.vector.reduce.add.v32i8(<32 x i8> undef)15; CHECK-NEXT: Cost Model: Found costs of 5 for: %V64i8 = call i8 @llvm.vector.reduce.add.v64i8(<64 x i8> undef)16; CHECK-NEXT: Cost Model: Found costs of 2 for: %V2i16 = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> undef)17; CHECK-NEXT: Cost Model: Found costs of 2 for: %V4i16 = call i16 @llvm.vector.reduce.add.v4i16(<4 x i16> undef)18; CHECK-NEXT: Cost Model: Found costs of 2 for: %V8i16 = call i16 @llvm.vector.reduce.add.v8i16(<8 x i16> undef)19; CHECK-NEXT: Cost Model: Found costs of 3 for: %V16i16 = call i16 @llvm.vector.reduce.add.v16i16(<16 x i16> undef)20; CHECK-NEXT: Cost Model: Found costs of 2 for: %V2i32 = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> undef)21; CHECK-NEXT: Cost Model: Found costs of 2 for: %V4i32 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> undef)22; CHECK-NEXT: Cost Model: Found costs of 3 for: %V8i32 = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> undef)23; CHECK-NEXT: Cost Model: Found costs of 2 for: %V2i64 = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> undef)24; CHECK-NEXT: Cost Model: Found costs of 3 for: %V4i64 = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> undef)25; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void26;27 %V1i8 = call i8 @llvm.vector.reduce.add.v1i8(<1 x i8> undef)28 %V2i8 = call i8 @llvm.vector.reduce.add.v2i8(<2 x i8> undef)29 %V3i8 = call i8 @llvm.vector.reduce.add.v3i8(<3 x i8> undef)30 %V4i8 = call i8 @llvm.vector.reduce.add.v4i8(<4 x i8> undef)31 %V8i8 = call i8 @llvm.vector.reduce.add.v8i8(<8 x i8> undef)32 %V16i8 = call i8 @llvm.vector.reduce.add.v16i8(<16 x i8> undef)33 %V32i8 = call i8 @llvm.vector.reduce.add.v32i8(<32 x i8> undef)34 %V64i8 = call i8 @llvm.vector.reduce.add.v64i8(<64 x i8> undef)35 %V2i16 = call i16 @llvm.vector.reduce.add.v2i16(<2 x i16> undef)36 %V4i16 = call i16 @llvm.vector.reduce.add.v4i16(<4 x i16> undef)37 %V8i16 = call i16 @llvm.vector.reduce.add.v8i16(<8 x i16> undef)38 %V16i16 = call i16 @llvm.vector.reduce.add.v16i16(<16 x i16> undef)39 %V2i32 = call i32 @llvm.vector.reduce.add.v2i32(<2 x i32> undef)40 %V4i32 = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> undef)41 %V8i32 = call i32 @llvm.vector.reduce.add.v8i32(<8 x i32> undef)42 %V2i64 = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> undef)43 %V4i64 = call i64 @llvm.vector.reduce.add.v4i64(<4 x i64> undef)44 ret void45}46