269 lines · plain
1; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py2; RUN: opt -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mtriple=thumbv8m.main < %s | FileCheck %s --check-prefix=V8M3; RUN: opt -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mtriple=armv8a-linux-gnueabihf < %s | FileCheck %s --check-prefix=NEON4; RUN: opt -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mtriple=armv8.1m.main -mattr=+mve < %s | FileCheck %s --check-prefix=MVE5 6target datalayout = "e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64"7 8declare i64 @llvm.sadd.sat.i64(i64, i64)9declare <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64>, <2 x i64>)10declare <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64>, <4 x i64>)11declare <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64>, <8 x i64>)12 13declare i32 @llvm.sadd.sat.i32(i32, i32)14declare <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32>, <2 x i32>)15declare <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32>, <4 x i32>)16declare <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32>, <8 x i32>)17declare <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32>, <16 x i32>)18 19declare i16 @llvm.sadd.sat.i16(i16, i16)20declare <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16>, <2 x i16>)21declare <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16>, <4 x i16>)22declare <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16>, <8 x i16>)23declare <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16>, <16 x i16>)24declare <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16>, <32 x i16>)25 26declare i8 @llvm.sadd.sat.i8(i8, i8)27declare <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8>, <2 x i8>)28declare <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8>, <4 x i8>)29declare <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8>, <8 x i8>)30declare <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8>, <16 x i8>)31declare <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8>, <32 x i8>)32declare <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8>, <64 x i8>)33 34define i32 @add(i32 %arg) {35; V8M-LABEL: 'add'36; V8M-NEXT: Cost Model: Found costs of RThru:8 CodeSize:10 Lat:8 SizeLat:8 for: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)37; V8M-NEXT: Cost Model: Found costs of RThru:30 CodeSize:34 Lat:30 SizeLat:30 for: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)38; V8M-NEXT: Cost Model: Found costs of RThru:60 CodeSize:68 Lat:60 SizeLat:60 for: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)39; V8M-NEXT: Cost Model: Found costs of RThru:120 CodeSize:136 Lat:120 SizeLat:120 for: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)40; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)41; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)42; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)43; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)44; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)45; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)46; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)47; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)48; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)49; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)50; V8M-NEXT: Cost Model: Found costs of RThru:256 CodeSize:320 Lat:256 SizeLat:256 for: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)51; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)52; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)53; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)54; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)55; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)56; V8M-NEXT: Cost Model: Found costs of RThru:256 CodeSize:320 Lat:256 SizeLat:256 for: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)57; V8M-NEXT: Cost Model: Found costs of RThru:512 CodeSize:640 Lat:512 SizeLat:512 for: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)58; V8M-NEXT: Cost Model: Found costs of 1 for: ret i32 undef59;60; NEON-LABEL: 'add'61; NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:7 Lat:7 SizeLat:7 for: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)62; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)63; NEON-NEXT: Cost Model: Found costs of 4 for: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)64; NEON-NEXT: Cost Model: Found costs of 8 for: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)65; NEON-NEXT: Cost Model: Found costs of 1 for: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)66; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)67; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)68; NEON-NEXT: Cost Model: Found costs of 4 for: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)69; NEON-NEXT: Cost Model: Found costs of 8 for: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)70; NEON-NEXT: Cost Model: Found costs of 2 for: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)71; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)72; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)73; NEON-NEXT: Cost Model: Found costs of 1 for: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)74; NEON-NEXT: Cost Model: Found costs of 4 for: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)75; NEON-NEXT: Cost Model: Found costs of 8 for: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)76; NEON-NEXT: Cost Model: Found costs of 2 for: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)77; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)78; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)79; NEON-NEXT: Cost Model: Found costs of 1 for: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)80; NEON-NEXT: Cost Model: Found costs of 1 for: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)81; NEON-NEXT: Cost Model: Found costs of 4 for: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)82; NEON-NEXT: Cost Model: Found costs of 8 for: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)83; NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef84;85; MVE-LABEL: 'add'86; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:10 Lat:8 SizeLat:8 for: %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)87; MVE-NEXT: Cost Model: Found costs of RThru:166 CodeSize:112 Lat:166 SizeLat:166 for: %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)88; MVE-NEXT: Cost Model: Found costs of RThru:298 CodeSize:185 Lat:298 SizeLat:298 for: %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)89; MVE-NEXT: Cost Model: Found costs of RThru:594 CodeSize:369 Lat:594 SizeLat:594 for: %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)90; MVE-NEXT: Cost Model: Found costs of 1 for: %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)91; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)92; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)93; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)94; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)95; MVE-NEXT: Cost Model: Found costs of 2 for: %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)96; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)97; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)98; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)99; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)100; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)101; MVE-NEXT: Cost Model: Found costs of 2 for: %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)102; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)103; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)104; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)105; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)106; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)107; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)108; MVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef109;110 %I64 = call i64 @llvm.sadd.sat.i64(i64 undef, i64 undef)111 %V2I64 = call <2 x i64> @llvm.sadd.sat.v2i64(<2 x i64> undef, <2 x i64> undef)112 %V4I64 = call <4 x i64> @llvm.sadd.sat.v4i64(<4 x i64> undef, <4 x i64> undef)113 %V8I64 = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> undef, <8 x i64> undef)114 115 %I32 = call i32 @llvm.sadd.sat.i32(i32 undef, i32 undef)116 %V2I32 = call <2 x i32> @llvm.sadd.sat.v2i32(<2 x i32> undef, <2 x i32> undef)117 %V4I32 = call <4 x i32> @llvm.sadd.sat.v4i32(<4 x i32> undef, <4 x i32> undef)118 %V8I32 = call <8 x i32> @llvm.sadd.sat.v8i32(<8 x i32> undef, <8 x i32> undef)119 %V16I32 = call <16 x i32> @llvm.sadd.sat.v16i32(<16 x i32> undef, <16 x i32> undef)120 121 %I16 = call i16 @llvm.sadd.sat.i16(i16 undef, i16 undef)122 %V2I16 = call <2 x i16> @llvm.sadd.sat.v2i16(<2 x i16> undef, <2 x i16> undef)123 %V4I16 = call <4 x i16> @llvm.sadd.sat.v4i16(<4 x i16> undef, <4 x i16> undef)124 %V8I16 = call <8 x i16> @llvm.sadd.sat.v8i16(<8 x i16> undef, <8 x i16> undef)125 %V16I16 = call <16 x i16> @llvm.sadd.sat.v16i16(<16 x i16> undef, <16 x i16> undef)126 %V32I16 = call <32 x i16> @llvm.sadd.sat.v32i16(<32 x i16> undef, <32 x i16> undef)127 128 %I8 = call i8 @llvm.sadd.sat.i8(i8 undef, i8 undef)129 %V2I8 = call <2 x i8> @llvm.sadd.sat.v2i8(<2 x i8> undef, <2 x i8> undef)130 %V4I8 = call <4 x i8> @llvm.sadd.sat.v4i8(<4 x i8> undef, <4 x i8> undef)131 %V8I8 = call <8 x i8> @llvm.sadd.sat.v8i8(<8 x i8> undef, <8 x i8> undef)132 %V16I8 = call <16 x i8> @llvm.sadd.sat.v16i8(<16 x i8> undef, <16 x i8> undef)133 %V32I8 = call <32 x i8> @llvm.sadd.sat.v32i8(<32 x i8> undef, <32 x i8> undef)134 %V64I8 = call <64 x i8> @llvm.sadd.sat.v64i8(<64 x i8> undef, <64 x i8> undef)135 136 ret i32 undef137}138 139declare i64 @llvm.ssub.sat.i64(i64, i64)140declare <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64>, <2 x i64>)141declare <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64>, <4 x i64>)142declare <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64>, <8 x i64>)143 144declare i32 @llvm.ssub.sat.i32(i32, i32)145declare <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32>, <2 x i32>)146declare <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32>, <4 x i32>)147declare <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32>, <8 x i32>)148declare <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32>, <16 x i32>)149 150declare i16 @llvm.ssub.sat.i16(i16, i16)151declare <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16>, <2 x i16>)152declare <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16>, <4 x i16>)153declare <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16>, <8 x i16>)154declare <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16>, <16 x i16>)155declare <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16>, <32 x i16>)156 157declare i8 @llvm.ssub.sat.i8(i8, i8)158declare <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8>, <2 x i8>)159declare <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8>, <4 x i8>)160declare <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8>, <8 x i8>)161declare <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8>, <16 x i8>)162declare <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8>, <32 x i8>)163declare <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8>, <64 x i8>)164 165define i32 @sub(i32 %arg) {166; V8M-LABEL: 'sub'167; V8M-NEXT: Cost Model: Found costs of RThru:8 CodeSize:10 Lat:8 SizeLat:8 for: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)168; V8M-NEXT: Cost Model: Found costs of RThru:30 CodeSize:34 Lat:30 SizeLat:30 for: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)169; V8M-NEXT: Cost Model: Found costs of RThru:60 CodeSize:68 Lat:60 SizeLat:60 for: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)170; V8M-NEXT: Cost Model: Found costs of RThru:120 CodeSize:136 Lat:120 SizeLat:120 for: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)171; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)172; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)173; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)174; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)175; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)176; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)177; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)178; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)179; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)180; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)181; V8M-NEXT: Cost Model: Found costs of RThru:256 CodeSize:320 Lat:256 SizeLat:256 for: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)182; V8M-NEXT: Cost Model: Found costs of RThru:5 CodeSize:7 Lat:5 SizeLat:5 for: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)183; V8M-NEXT: Cost Model: Found costs of RThru:16 CodeSize:20 Lat:16 SizeLat:16 for: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)184; V8M-NEXT: Cost Model: Found costs of RThru:32 CodeSize:40 Lat:32 SizeLat:32 for: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)185; V8M-NEXT: Cost Model: Found costs of RThru:64 CodeSize:80 Lat:64 SizeLat:64 for: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)186; V8M-NEXT: Cost Model: Found costs of RThru:128 CodeSize:160 Lat:128 SizeLat:128 for: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)187; V8M-NEXT: Cost Model: Found costs of RThru:256 CodeSize:320 Lat:256 SizeLat:256 for: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)188; V8M-NEXT: Cost Model: Found costs of RThru:512 CodeSize:640 Lat:512 SizeLat:512 for: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)189; V8M-NEXT: Cost Model: Found costs of 1 for: ret i32 undef190;191; NEON-LABEL: 'sub'192; NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:7 Lat:7 SizeLat:7 for: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)193; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)194; NEON-NEXT: Cost Model: Found costs of 4 for: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)195; NEON-NEXT: Cost Model: Found costs of 8 for: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)196; NEON-NEXT: Cost Model: Found costs of 1 for: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)197; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)198; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)199; NEON-NEXT: Cost Model: Found costs of 4 for: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)200; NEON-NEXT: Cost Model: Found costs of 8 for: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)201; NEON-NEXT: Cost Model: Found costs of 2 for: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)202; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)203; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)204; NEON-NEXT: Cost Model: Found costs of 1 for: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)205; NEON-NEXT: Cost Model: Found costs of 4 for: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)206; NEON-NEXT: Cost Model: Found costs of 8 for: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)207; NEON-NEXT: Cost Model: Found costs of 2 for: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)208; NEON-NEXT: Cost Model: Found costs of 1 for: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)209; NEON-NEXT: Cost Model: Found costs of 1 for: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)210; NEON-NEXT: Cost Model: Found costs of 1 for: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)211; NEON-NEXT: Cost Model: Found costs of 1 for: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)212; NEON-NEXT: Cost Model: Found costs of 4 for: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)213; NEON-NEXT: Cost Model: Found costs of 8 for: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)214; NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef215;216; MVE-LABEL: 'sub'217; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:10 Lat:8 SizeLat:8 for: %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)218; MVE-NEXT: Cost Model: Found costs of RThru:166 CodeSize:112 Lat:166 SizeLat:166 for: %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)219; MVE-NEXT: Cost Model: Found costs of RThru:298 CodeSize:185 Lat:298 SizeLat:298 for: %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)220; MVE-NEXT: Cost Model: Found costs of RThru:594 CodeSize:369 Lat:594 SizeLat:594 for: %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)221; MVE-NEXT: Cost Model: Found costs of 1 for: %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)222; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)223; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)224; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)225; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)226; MVE-NEXT: Cost Model: Found costs of 2 for: %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)227; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)228; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)229; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)230; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)231; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)232; MVE-NEXT: Cost Model: Found costs of 2 for: %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)233; MVE-NEXT: Cost Model: Found costs of RThru:100 CodeSize:70 Lat:100 SizeLat:100 for: %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)234; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)235; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)236; MVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:2 SizeLat:2 for: %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)237; MVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)238; MVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:4 Lat:8 SizeLat:8 for: %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)239; MVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef240;241 %I64 = call i64 @llvm.ssub.sat.i64(i64 undef, i64 undef)242 %V2I64 = call <2 x i64> @llvm.ssub.sat.v2i64(<2 x i64> undef, <2 x i64> undef)243 %V4I64 = call <4 x i64> @llvm.ssub.sat.v4i64(<4 x i64> undef, <4 x i64> undef)244 %V8I64 = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> undef, <8 x i64> undef)245 246 %I32 = call i32 @llvm.ssub.sat.i32(i32 undef, i32 undef)247 %V2I32 = call <2 x i32> @llvm.ssub.sat.v2i32(<2 x i32> undef, <2 x i32> undef)248 %V4I32 = call <4 x i32> @llvm.ssub.sat.v4i32(<4 x i32> undef, <4 x i32> undef)249 %V8I32 = call <8 x i32> @llvm.ssub.sat.v8i32(<8 x i32> undef, <8 x i32> undef)250 %V16I32 = call <16 x i32> @llvm.ssub.sat.v16i32(<16 x i32> undef, <16 x i32> undef)251 252 %I16 = call i16 @llvm.ssub.sat.i16(i16 undef, i16 undef)253 %V2I16 = call <2 x i16> @llvm.ssub.sat.v2i16(<2 x i16> undef, <2 x i16> undef)254 %V4I16 = call <4 x i16> @llvm.ssub.sat.v4i16(<4 x i16> undef, <4 x i16> undef)255 %V8I16 = call <8 x i16> @llvm.ssub.sat.v8i16(<8 x i16> undef, <8 x i16> undef)256 %V16I16 = call <16 x i16> @llvm.ssub.sat.v16i16(<16 x i16> undef, <16 x i16> undef)257 %V32I16 = call <32 x i16> @llvm.ssub.sat.v32i16(<32 x i16> undef, <32 x i16> undef)258 259 %I8 = call i8 @llvm.ssub.sat.i8(i8 undef, i8 undef)260 %V2I8 = call <2 x i8> @llvm.ssub.sat.v2i8(<2 x i8> undef, <2 x i8> undef)261 %V4I8 = call <4 x i8> @llvm.ssub.sat.v4i8(<4 x i8> undef, <4 x i8> undef)262 %V8I8 = call <8 x i8> @llvm.ssub.sat.v8i8(<8 x i8> undef, <8 x i8> undef)263 %V16I8 = call <16 x i8> @llvm.ssub.sat.v16i8(<16 x i8> undef, <16 x i8> undef)264 %V32I8 = call <32 x i8> @llvm.ssub.sat.v32i8(<32 x i8> undef, <32 x i8> undef)265 %V64I8 = call <64 x i8> @llvm.ssub.sat.v64i8(<64 x i8> undef, <64 x i8> undef)266 267 ret i32 undef268}269