294 lines · plain
1; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py2; RUN: opt < %s -passes="print<cost-model>" -cost-kind=all 2>&1 -disable-output -mtriple=aarch64-linux-gnu -mattr=+sve | FileCheck %s3 4target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"5 6define void @fixed() {7; CHECK-LABEL: 'fixed'8; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:8 Lat:12 SizeLat:12 for: %v2i8 = call <2 x i8> @llvm.masked.load.v2i8.p0(ptr align 8 undef, <2 x i1> undef, <2 x i8> undef)9; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:16 Lat:24 SizeLat:24 for: %v4i8 = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr align 8 undef, <4 x i1> undef, <4 x i8> undef)10; CHECK-NEXT: Cost Model: Found costs of RThru:40 CodeSize:32 Lat:48 SizeLat:48 for: %v8i8 = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr align 8 undef, <8 x i1> undef, <8 x i8> undef)11; CHECK-NEXT: Cost Model: Found costs of RThru:80 CodeSize:64 Lat:96 SizeLat:96 for: %v16i8 = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr align 8 undef, <16 x i1> undef, <16 x i8> undef)12; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:8 Lat:12 SizeLat:12 for: %v2i16 = call <2 x i16> @llvm.masked.load.v2i16.p0(ptr align 8 undef, <2 x i1> undef, <2 x i16> undef)13; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:16 Lat:24 SizeLat:24 for: %v4i16 = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr align 8 undef, <4 x i1> undef, <4 x i16> undef)14; CHECK-NEXT: Cost Model: Found costs of RThru:40 CodeSize:32 Lat:48 SizeLat:48 for: %v8i16 = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr align 8 undef, <8 x i1> undef, <8 x i16> undef)15; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:8 Lat:12 SizeLat:12 for: %v2i32 = call <2 x i32> @llvm.masked.load.v2i32.p0(ptr align 8 undef, <2 x i1> undef, <2 x i32> undef)16; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:16 Lat:24 SizeLat:24 for: %v4i32 = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr align 8 undef, <4 x i1> undef, <4 x i32> undef)17; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:8 Lat:12 SizeLat:12 for: %v2i64 = call <2 x i64> @llvm.masked.load.v2i64.p0(ptr align 8 undef, <2 x i1> undef, <2 x i64> undef)18; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:7 Lat:10 SizeLat:10 for: %v2f16 = call <2 x half> @llvm.masked.load.v2f16.p0(ptr align 8 undef, <2 x i1> undef, <2 x half> undef)19; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:15 Lat:22 SizeLat:22 for: %v4f16 = call <4 x half> @llvm.masked.load.v4f16.p0(ptr align 8 undef, <4 x i1> undef, <4 x half> undef)20; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:31 Lat:46 SizeLat:46 for: %v8f16 = call <8 x half> @llvm.masked.load.v8f16.p0(ptr align 8 undef, <8 x i1> undef, <8 x half> undef)21; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:7 Lat:10 SizeLat:10 for: %v2f32 = call <2 x float> @llvm.masked.load.v2f32.p0(ptr align 8 undef, <2 x i1> undef, <2 x float> undef)22; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:15 Lat:22 SizeLat:22 for: %v4f32 = call <4 x float> @llvm.masked.load.v4f32.p0(ptr align 8 undef, <4 x i1> undef, <4 x float> undef)23; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:7 Lat:10 SizeLat:10 for: %v2f64 = call <2 x double> @llvm.masked.load.v2f64.p0(ptr align 8 undef, <2 x i1> undef, <2 x double> undef)24; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:16 Lat:24 SizeLat:24 for: %v4i64 = call <4 x i64> @llvm.masked.load.v4i64.p0(ptr align 8 undef, <4 x i1> undef, <4 x i64> undef)25; CHECK-NEXT: Cost Model: Found costs of RThru:152 CodeSize:124 Lat:184 SizeLat:184 for: %v32f16 = call <32 x half> @llvm.masked.load.v32f16.p0(ptr align 8 undef, <32 x i1> undef, <32 x half> undef)26; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void27;28entry:29 ; Legal fixed-width integer types30 %v2i8 = call <2 x i8> @llvm.masked.load.v2i8.p0(ptr undef, i32 8, <2 x i1> undef, <2 x i8> undef)31 %v4i8 = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr undef, i32 8, <4 x i1> undef, <4 x i8> undef)32 %v8i8 = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr undef, i32 8, <8 x i1> undef, <8 x i8> undef)33 %v16i8 = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr undef, i32 8, <16 x i1> undef, <16 x i8> undef)34 %v2i16 = call <2 x i16> @llvm.masked.load.v2i16.p0(ptr undef, i32 8, <2 x i1> undef, <2 x i16> undef)35 %v4i16 = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr undef, i32 8, <4 x i1> undef, <4 x i16> undef)36 %v8i16 = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr undef, i32 8, <8 x i1> undef, <8 x i16> undef)37 %v2i32 = call <2 x i32> @llvm.masked.load.v2i32.p0(ptr undef, i32 8, <2 x i1> undef, <2 x i32> undef)38 %v4i32 = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr undef, i32 8, <4 x i1> undef, <4 x i32> undef)39 %v2i64 = call <2 x i64> @llvm.masked.load.v2i64.p0(ptr undef, i32 8, <2 x i1> undef, <2 x i64> undef)40 41 ; Legal fixed-width floating point types42 %v2f16 = call <2 x half> @llvm.masked.load.v2f16.p0(ptr undef, i32 8, <2 x i1> undef, <2 x half> undef)43 %v4f16 = call <4 x half> @llvm.masked.load.v4f16.p0(ptr undef, i32 8, <4 x i1> undef, <4 x half> undef)44 %v8f16 = call <8 x half> @llvm.masked.load.v8f16.p0(ptr undef, i32 8, <8 x i1> undef, <8 x half> undef)45 %v2f32 = call <2 x float> @llvm.masked.load.v2f32.p0(ptr undef, i32 8, <2 x i1> undef, <2 x float> undef)46 %v4f32 = call <4 x float> @llvm.masked.load.v4f32.p0(ptr undef, i32 8, <4 x i1> undef, <4 x float> undef)47 %v2f64 = call <2 x double> @llvm.masked.load.v2f64.p0(ptr undef, i32 8, <2 x i1> undef, <2 x double> undef)48 49 ; A couple of examples of illegal fixed-width types50 %v4i64 = call <4 x i64> @llvm.masked.load.v4i64.p0(ptr undef, i32 8, <4 x i1> undef, <4 x i64> undef)51 %v32f16 = call <32 x half> @llvm.masked.load.v32f16.p0(ptr undef, i32 8, <32 x i1> undef, <32 x half> undef)52 53 ret void54}55 56 57define void @scalable() {58; CHECK-LABEL: 'scalable'59; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)60; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)61; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)62; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)63; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)64; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)65; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)66; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)67; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)68; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2i64 = call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i64> undef)69; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2f16 = call <vscale x 2 x half> @llvm.masked.load.nxv2f16.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x half> undef)70; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4f16 = call <vscale x 4 x half> @llvm.masked.load.nxv4f16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x half> undef)71; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv8f16 = call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x half> undef)72; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2f32 = call <vscale x 2 x float> @llvm.masked.load.nxv2f32.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x float> undef)73; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4f32 = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x float> undef)74; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2f64 = call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x double> undef)75; CHECK-NEXT: Cost Model: Found costs of Invalid for: %nxv1i64 = call <vscale x 1 x i64> @llvm.masked.load.nxv1i64.p0(ptr align 8 undef, <vscale x 1 x i1> undef, <vscale x 1 x i64> undef)76; CHECK-NEXT: Cost Model: Found costs of 2 for: %nxv4i64 = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i64> undef)77; CHECK-NEXT: Cost Model: Found costs of 4 for: %nxv32f16 = call <vscale x 32 x half> @llvm.masked.load.nxv32f16.p0(ptr align 8 undef, <vscale x 32 x i1> undef, <vscale x 32 x half> undef)78; CHECK-NEXT: Cost Model: Found costs of Invalid for: %nxv4i1 = call <vscale x 4 x i1> @llvm.masked.load.nxv4i1.p0(ptr align 16 undef, <vscale x 4 x i1> undef, <vscale x 4 x i1> undef)79; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void80;81entry:82 ; Legal scalable integer types83 %nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)84 %nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)85 %nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)86 %nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)87 %nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)88 %nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)89 %nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)90 %nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)91 %nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)92 %nxv2i64 = call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i64> undef)93 94 ; Legal scalable floating point types95 %nxv2f16 = call <vscale x 2 x half> @llvm.masked.load.nxv2f16.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x half> undef)96 %nxv4f16 = call <vscale x 4 x half> @llvm.masked.load.nxv4f16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x half> undef)97 %nxv8f16 = call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x half> undef)98 %nxv2f32 = call <vscale x 2 x float> @llvm.masked.load.nxv2f32.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x float> undef)99 %nxv4f32 = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x float> undef)100 %nxv2f64 = call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x double> undef)101 102 ; A couple of examples of illegal scalable types103 %nxv1i64 = call <vscale x 1 x i64> @llvm.masked.load.nxv1i64.p0(ptr undef, i32 8, <vscale x 1 x i1> undef, <vscale x 1 x i64> undef)104 %nxv4i64 = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i64> undef)105 %nxv32f16 = call <vscale x 32 x half> @llvm.masked.load.nxv32f16.p0(ptr undef, i32 8, <vscale x 32 x i1> undef, <vscale x 32 x half> undef)106 107 ; Types that are legal, but for which we have no masked load/store lowering108 %nxv4i1 = call <vscale x 4 x i1> @llvm.masked.load.nxv4i1.p0(ptr undef, i32 16, <vscale x 4 x i1> undef, <vscale x 4 x i1> undef)109 110 ret void111}112 113 114define void @scalable_ext_loads() {115; CHECK-LABEL: 'scalable_ext_loads'116; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)117; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv16i8to16 = zext <vscale x 16 x i8> %load.nxv16i8 to <vscale x 16 x i16>118; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv16i8.2 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)119; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv16i8to32 = zext <vscale x 16 x i8> %load.nxv16i8.2 to <vscale x 16 x i32>120; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv16i8.3 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)121; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv16i8to64 = zext <vscale x 16 x i8> %load.nxv16i8.3 to <vscale x 16 x i64>122; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)123; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv8i8to16 = zext <vscale x 8 x i8> %load.nxv8i8 to <vscale x 8 x i16>124; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv8i8.2 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)125; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv8i8to32 = zext <vscale x 8 x i8> %load.nxv8i8.2 to <vscale x 8 x i32>126; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv8i8.3 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)127; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv8i8to64 = zext <vscale x 8 x i8> %load.nxv8i8.3 to <vscale x 8 x i64>128; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)129; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv4i8to32 = zext <vscale x 4 x i8> %load.nxv4i8 to <vscale x 4 x i32>130; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv4i8.2 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)131; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv4i8to64 = zext <vscale x 4 x i8> %load.nxv4i8.2 to <vscale x 4 x i64>132; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)133; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv2i8to64 = zext <vscale x 2 x i8> %load.nxv2i8 to <vscale x 2 x i64>134; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)135; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv8i16to32 = zext <vscale x 8 x i16> %load.nxv8i16 to <vscale x 8 x i32>136; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv8i16.2 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)137; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv8i16to64 = zext <vscale x 8 x i16> %load.nxv8i16.2 to <vscale x 8 x i64>138; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)139; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv4i16to32 = zext <vscale x 4 x i16> %load.nxv4i16 to <vscale x 4 x i32>140; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv4i16.2 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)141; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv4i16to64 = zext <vscale x 4 x i16> %load.nxv4i16.2 to <vscale x 4 x i64>142; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)143; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv2i16to64 = zext <vscale x 2 x i16> %load.nxv2i16 to <vscale x 2 x i64>144; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)145; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext.nxv4i32to64 = zext <vscale x 4 x i32> %load.nxv4i32 to <vscale x 4 x i64>146; CHECK-NEXT: Cost Model: Found costs of 1 for: %load.nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)147; CHECK-NEXT: Cost Model: Found costs of 0 for: %zext.nxv2i32to64 = zext <vscale x 2 x i32> %load.nxv2i32 to <vscale x 2 x i64>148; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)149; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv16i8to16 = sext <vscale x 16 x i8> %load2.nxv16i8 to <vscale x 16 x i16>150; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv16i8.2 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)151; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv16i8to32 = sext <vscale x 16 x i8> %load2.nxv16i8.2 to <vscale x 16 x i32>152; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv16i8.3 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 8 undef, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)153; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv16i8to64 = sext <vscale x 16 x i8> %load2.nxv16i8.3 to <vscale x 16 x i64>154; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)155; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv8i8to16 = sext <vscale x 8 x i8> %load2.nxv8i8 to <vscale x 8 x i16>156; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv8i8.2 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)157; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv8i8to32 = sext <vscale x 8 x i8> %load2.nxv8i8.2 to <vscale x 8 x i32>158; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv8i8.3 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)159; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv8i8to64 = sext <vscale x 8 x i8> %load2.nxv8i8.3 to <vscale x 8 x i64>160; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)161; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv4i8to32 = sext <vscale x 4 x i8> %load2.nxv4i8 to <vscale x 4 x i32>162; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv4i8.2 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)163; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv4i8to64 = sext <vscale x 4 x i8> %load2.nxv4i8.2 to <vscale x 4 x i64>164; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)165; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv2i8to64 = sext <vscale x 2 x i8> %load2.nxv2i8 to <vscale x 2 x i64>166; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)167; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv8i16to32 = sext <vscale x 8 x i16> %load2.nxv8i16 to <vscale x 8 x i32>168; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv8i16.2 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 8 undef, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)169; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv8i16to64 = sext <vscale x 8 x i16> %load2.nxv8i16.2 to <vscale x 8 x i64>170; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)171; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv4i16to32 = sext <vscale x 4 x i16> %load2.nxv4i16 to <vscale x 4 x i32>172; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv4i16.2 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)173; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv4i16to64 = sext <vscale x 4 x i16> %load2.nxv4i16.2 to <vscale x 4 x i64>174; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)175; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv2i16to64 = sext <vscale x 2 x i16> %load2.nxv2i16 to <vscale x 2 x i64>176; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 8 undef, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)177; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext.nxv4i32to64 = sext <vscale x 4 x i32> %load2.nxv4i32 to <vscale x 4 x i64>178; CHECK-NEXT: Cost Model: Found costs of 1 for: %load2.nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 8 undef, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)179; CHECK-NEXT: Cost Model: Found costs of 0 for: %sext.nxv2i32to64 = sext <vscale x 2 x i32> %load2.nxv2i32 to <vscale x 2 x i64>180; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void181;182 183 184 %load.nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)185 %zext.nxv16i8to16 = zext <vscale x 16 x i8> %load.nxv16i8 to <vscale x 16 x i16>186 %load.nxv16i8.2 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)187 %zext.nxv16i8to32 = zext <vscale x 16 x i8> %load.nxv16i8.2 to <vscale x 16 x i32>188 %load.nxv16i8.3 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)189 %zext.nxv16i8to64 = zext <vscale x 16 x i8> %load.nxv16i8.3 to <vscale x 16 x i64>190 %load.nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)191 %zext.nxv8i8to16 = zext <vscale x 8 x i8> %load.nxv8i8 to <vscale x 8 x i16>192 %load.nxv8i8.2 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)193 %zext.nxv8i8to32 = zext <vscale x 8 x i8> %load.nxv8i8.2 to <vscale x 8 x i32>194 %load.nxv8i8.3 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)195 %zext.nxv8i8to64 = zext <vscale x 8 x i8> %load.nxv8i8.3 to <vscale x 8 x i64>196 %load.nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)197 %zext.nxv4i8to32 = zext <vscale x 4 x i8> %load.nxv4i8 to <vscale x 4 x i32>198 %load.nxv4i8.2 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)199 %zext.nxv4i8to64 = zext <vscale x 4 x i8> %load.nxv4i8.2 to <vscale x 4 x i64>200 %load.nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)201 %zext.nxv2i8to64 = zext <vscale x 2 x i8> %load.nxv2i8 to <vscale x 2 x i64>202 %load.nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)203 %zext.nxv8i16to32 = zext <vscale x 8 x i16> %load.nxv8i16 to <vscale x 8 x i32>204 %load.nxv8i16.2 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)205 %zext.nxv8i16to64 = zext <vscale x 8 x i16> %load.nxv8i16.2 to <vscale x 8 x i64>206 %load.nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)207 %zext.nxv4i16to32 = zext <vscale x 4 x i16> %load.nxv4i16 to <vscale x 4 x i32>208 %load.nxv4i16.2 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)209 %zext.nxv4i16to64 = zext <vscale x 4 x i16> %load.nxv4i16.2 to <vscale x 4 x i64>210 %load.nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)211 %zext.nxv2i16to64 = zext <vscale x 2 x i16> %load.nxv2i16 to <vscale x 2 x i64>212 %load.nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)213 %zext.nxv4i32to64 = zext <vscale x 4 x i32> %load.nxv4i32 to <vscale x 4 x i64>214 %load.nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)215 %zext.nxv2i32to64 = zext <vscale x 2 x i32> %load.nxv2i32 to <vscale x 2 x i64>216 217 %load2.nxv16i8 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)218 %sext.nxv16i8to16 = sext <vscale x 16 x i8> %load2.nxv16i8 to <vscale x 16 x i16>219 %load2.nxv16i8.2 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)220 %sext.nxv16i8to32 = sext <vscale x 16 x i8> %load2.nxv16i8.2 to <vscale x 16 x i32>221 %load2.nxv16i8.3 = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr undef, i32 8, <vscale x 16 x i1> undef, <vscale x 16 x i8> undef)222 %sext.nxv16i8to64 = sext <vscale x 16 x i8> %load2.nxv16i8.3 to <vscale x 16 x i64>223 %load2.nxv8i8 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)224 %sext.nxv8i8to16 = sext <vscale x 8 x i8> %load2.nxv8i8 to <vscale x 8 x i16>225 %load2.nxv8i8.2 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)226 %sext.nxv8i8to32 = sext <vscale x 8 x i8> %load2.nxv8i8.2 to <vscale x 8 x i32>227 %load2.nxv8i8.3 = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i8> undef)228 %sext.nxv8i8to64 = sext <vscale x 8 x i8> %load2.nxv8i8.3 to <vscale x 8 x i64>229 %load2.nxv4i8 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)230 %sext.nxv4i8to32 = sext <vscale x 4 x i8> %load2.nxv4i8 to <vscale x 4 x i32>231 %load2.nxv4i8.2 = call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i8> undef)232 %sext.nxv4i8to64 = sext <vscale x 4 x i8> %load2.nxv4i8.2 to <vscale x 4 x i64>233 %load2.nxv2i8 = call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i8> undef)234 %sext.nxv2i8to64 = sext <vscale x 2 x i8> %load2.nxv2i8 to <vscale x 2 x i64>235 %load2.nxv8i16 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)236 %sext.nxv8i16to32 = sext <vscale x 8 x i16> %load2.nxv8i16 to <vscale x 8 x i32>237 %load2.nxv8i16.2 = call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr undef, i32 8, <vscale x 8 x i1> undef, <vscale x 8 x i16> undef)238 %sext.nxv8i16to64 = sext <vscale x 8 x i16> %load2.nxv8i16.2 to <vscale x 8 x i64>239 %load2.nxv4i16 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)240 %sext.nxv4i16to32 = sext <vscale x 4 x i16> %load2.nxv4i16 to <vscale x 4 x i32>241 %load2.nxv4i16.2 = call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i16> undef)242 %sext.nxv4i16to64 = sext <vscale x 4 x i16> %load2.nxv4i16.2 to <vscale x 4 x i64>243 %load2.nxv2i16 = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i16> undef)244 %sext.nxv2i16to64 = sext <vscale x 2 x i16> %load2.nxv2i16 to <vscale x 2 x i64>245 %load2.nxv4i32 = call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr undef, i32 8, <vscale x 4 x i1> undef, <vscale x 4 x i32> undef)246 %sext.nxv4i32to64 = sext <vscale x 4 x i32> %load2.nxv4i32 to <vscale x 4 x i64>247 %load2.nxv2i32 = call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr undef, i32 8, <vscale x 2 x i1> undef, <vscale x 2 x i32> undef)248 %sext.nxv2i32to64 = sext <vscale x 2 x i32> %load2.nxv2i32 to <vscale x 2 x i64>249 250 ret void251}252 253 254declare <2 x i8> @llvm.masked.load.v2i8.p0(ptr, i32, <2 x i1>, <2 x i8>)255declare <4 x i8> @llvm.masked.load.v4i8.p0(ptr, i32, <4 x i1>, <4 x i8>)256declare <8 x i8> @llvm.masked.load.v8i8.p0(ptr, i32, <8 x i1>, <8 x i8>)257declare <16 x i8> @llvm.masked.load.v16i8.p0(ptr, i32, <16 x i1>, <16 x i8>)258declare <2 x i16> @llvm.masked.load.v2i16.p0(ptr, i32, <2 x i1>, <2 x i16>)259declare <4 x i16> @llvm.masked.load.v4i16.p0(ptr, i32, <4 x i1>, <4 x i16>)260declare <8 x i16> @llvm.masked.load.v8i16.p0(ptr, i32, <8 x i1>, <8 x i16>)261declare <2 x i32> @llvm.masked.load.v2i32.p0(ptr, i32, <2 x i1>, <2 x i32>)262declare <4 x i32> @llvm.masked.load.v4i32.p0(ptr, i32, <4 x i1>, <4 x i32>)263declare <2 x i64> @llvm.masked.load.v2i64.p0(ptr, i32, <2 x i1>, <2 x i64>)264declare <4 x i64> @llvm.masked.load.v4i64.p0(ptr, i32, <4 x i1>, <4 x i64>)265declare <2 x half> @llvm.masked.load.v2f16.p0(ptr, i32, <2 x i1>, <2 x half>)266declare <4 x half> @llvm.masked.load.v4f16.p0(ptr, i32, <4 x i1>, <4 x half>)267declare <8 x half> @llvm.masked.load.v8f16.p0(ptr, i32, <8 x i1>, <8 x half>)268declare <32 x half> @llvm.masked.load.v32f16.p0(ptr, i32, <32 x i1>, <32 x half>)269declare <2 x float> @llvm.masked.load.v2f32.p0(ptr, i32, <2 x i1>, <2 x float>)270declare <4 x float> @llvm.masked.load.v4f32.p0(ptr, i32, <4 x i1>, <4 x float>)271declare <2 x double> @llvm.masked.load.v2f64.p0(ptr, i32, <2 x i1>, <2 x double>)272declare <vscale x 4 x i1> @llvm.masked.load.nxv4i1.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i1>)273 274 275declare <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i8>)276declare <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i8>)277declare <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr, i32, <vscale x 8 x i1>, <vscale x 8 x i8>)278declare <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr, i32, <vscale x 16 x i1>, <vscale x 16 x i8>)279declare <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i16>)280declare <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i16>)281declare <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr, i32, <vscale x 8 x i1>, <vscale x 8 x i16>)282declare <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i32>)283declare <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i32>)284declare <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x i64>)285declare <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x i64>)286declare <vscale x 1 x i64> @llvm.masked.load.nxv1i64.p0(ptr, i32, <vscale x 1 x i1>, <vscale x 1 x i64>)287declare <vscale x 2 x half> @llvm.masked.load.nxv2f16.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x half>)288declare <vscale x 4 x half> @llvm.masked.load.nxv4f16.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x half>)289declare <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr, i32, <vscale x 8 x i1>, <vscale x 8 x half>)290declare <vscale x 32 x half> @llvm.masked.load.nxv32f16.p0(ptr, i32, <vscale x 32 x i1>, <vscale x 32 x half>)291declare <vscale x 2 x float> @llvm.masked.load.nxv2f32.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x float>)292declare <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr, i32, <vscale x 4 x i1>, <vscale x 4 x float>)293declare <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr, i32, <vscale x 2 x i1>, <vscale x 2 x double>)294