1; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py 2; RUN: opt < %s -cost-model -analyze -cost-kind=size-latency -mtriple=x86_64-- -mattr=+sse2 | FileCheck %s --check-prefixes=SSE2 3; RUN: opt < %s -cost-model -analyze -cost-kind=size-latency -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX2 4 5define i32 @f32(i32 %arg) { 6; SSE2-LABEL: 'f32' 7; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %F32 = call float @llvm.maxnum.f32(float undef, float undef) 8; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2F32 = call <2 x float> @llvm.maxnum.v2f32(<2 x float> undef, <2 x float> undef) 9; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4F32 = call <4 x float> @llvm.maxnum.v4f32(<4 x float> undef, <4 x float> undef) 10; SSE2-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V8F32 = call <8 x float> @llvm.maxnum.v8f32(<8 x float> undef, <8 x float> undef) 11; SSE2-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V16F32 = call <16 x float> @llvm.maxnum.v16f32(<16 x float> undef, <16 x float> undef) 12; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef 13; 14; AVX2-LABEL: 'f32' 15; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %F32 = call float @llvm.maxnum.f32(float undef, float undef) 16; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2F32 = call <2 x float> @llvm.maxnum.v2f32(<2 x float> undef, <2 x float> undef) 17; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4F32 = call <4 x float> @llvm.maxnum.v4f32(<4 x float> undef, <4 x float> undef) 18; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V8F32 = call <8 x float> @llvm.maxnum.v8f32(<8 x float> undef, <8 x float> undef) 19; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V16F32 = call <16 x float> @llvm.maxnum.v16f32(<16 x float> undef, <16 x float> undef) 20; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef 21; 22 %F32 = call float @llvm.maxnum.f32(float undef, float undef) 23 %V2F32 = call <2 x float> @llvm.maxnum.v2f32(<2 x float> undef, <2 x float> undef) 24 %V4F32 = call <4 x float> @llvm.maxnum.v4f32(<4 x float> undef, <4 x float> undef) 25 %V8F32 = call <8 x float> @llvm.maxnum.v8f32(<8 x float> undef, <8 x float> undef) 26 %V16F32 = call <16 x float> @llvm.maxnum.v16f32(<16 x float> undef, <16 x float> undef) 27 ret i32 undef 28} 29 30define i32 @f64(i32 %arg) { 31; SSE2-LABEL: 'f64' 32; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %f64 = call double @llvm.maxnum.f64(double undef, double undef) 33; SSE2-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2f64 = call <2 x double> @llvm.maxnum.v2f64(<2 x double> undef, <2 x double> undef) 34; SSE2-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4f64 = call <4 x double> @llvm.maxnum.v4f64(<4 x double> undef, <4 x double> undef) 35; SSE2-NEXT: Cost Model: Found an estimated cost of 16 for instruction: %V8f64 = call <8 x double> @llvm.maxnum.v8f64(<8 x double> undef, <8 x double> undef) 36; SSE2-NEXT: Cost Model: Found an estimated cost of 32 for instruction: %V16f64 = call <16 x double> @llvm.maxnum.v16f64(<16 x double> undef, <16 x double> undef) 37; SSE2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef 38; 39; AVX2-LABEL: 'f64' 40; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %f64 = call double @llvm.maxnum.f64(double undef, double undef) 41; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2f64 = call <2 x double> @llvm.maxnum.v2f64(<2 x double> undef, <2 x double> undef) 42; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4f64 = call <4 x double> @llvm.maxnum.v4f64(<4 x double> undef, <4 x double> undef) 43; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8f64 = call <8 x double> @llvm.maxnum.v8f64(<8 x double> undef, <8 x double> undef) 44; AVX2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16f64 = call <16 x double> @llvm.maxnum.v16f64(<16 x double> undef, <16 x double> undef) 45; AVX2-NEXT: Cost Model: Found an estimated cost of 1 for instruction: ret i32 undef 46; 47 %f64 = call double @llvm.maxnum.f64(double undef, double undef) 48 %V2f64 = call <2 x double> @llvm.maxnum.v2f64(<2 x double> undef, <2 x double> undef) 49 %V4f64 = call <4 x double> @llvm.maxnum.v4f64(<4 x double> undef, <4 x double> undef) 50 %V8f64 = call <8 x double> @llvm.maxnum.v8f64(<8 x double> undef, <8 x double> undef) 51 %V16f64 = call <16 x double> @llvm.maxnum.v16f64(<16 x double> undef, <16 x double> undef) 52 ret i32 undef 53} 54 55declare float @llvm.maxnum.f32(float, float) 56declare <2 x float> @llvm.maxnum.v2f32(<2 x float>, <2 x float>) 57declare <4 x float> @llvm.maxnum.v4f32(<4 x float>, <4 x float>) 58declare <8 x float> @llvm.maxnum.v8f32(<8 x float>, <8 x float>) 59declare <16 x float> @llvm.maxnum.v16f32(<16 x float>, <16 x float>) 60 61declare double @llvm.maxnum.f64(double, double) 62declare <2 x double> @llvm.maxnum.v2f64(<2 x double>, <2 x double>) 63declare <4 x double> @llvm.maxnum.v4f64(<4 x double>, <4 x double>) 64declare <8 x double> @llvm.maxnum.v8f64(<8 x double>, <8 x double>) 65declare <16 x double> @llvm.maxnum.v16f64(<16 x double>, <16 x double>) 66