| 1 |
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py |
| 2 |
// REQUIRES: riscv-registered-target |
| 3 |
// RUN: %clang_cc1 -triple riscv64 -target-feature +zve64x \ |
| 4 |
// RUN: -target-feature +xandesvdot -disable-O0-optnone \ |
| 5 |
// RUN: -emit-llvm %s -o - | opt -S -passes=mem2reg | \ |
| 6 |
// RUN: FileCheck --check-prefix=CHECK-RV64 %s |
| 7 |
|
| 8 |
#include <andes_vector.h> |
| 9 |
|
| 10 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32mf2_tu( |
| 11 |
// CHECK-RV64-NEXT: entry: |
| 12 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.nds.vd4dotsu.nxv1i32.nxv4i8.nxv4i8.i64(<vscale x 1 x i32> [[VD:%.*]], <vscale x 4 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 13 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i32> [[TMP0]] |
| 14 |
// |
| 15 |
vint32mf2_t test_nds_vd4dotsu_vv_i32mf2_tu(vint32mf2_t vd, vint8mf2_t vs1, vuint8mf2_t vs2, size_t vl) { |
| 16 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 17 |
} |
| 18 |
|
| 19 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m1_tu( |
| 20 |
// CHECK-RV64-NEXT: entry: |
| 21 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.nds.vd4dotsu.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 22 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i32> [[TMP0]] |
| 23 |
// |
| 24 |
vint32m1_t test_nds_vd4dotsu_vv_i32m1_tu(vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) { |
| 25 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 26 |
} |
| 27 |
|
| 28 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m2_tu( |
| 29 |
// CHECK-RV64-NEXT: entry: |
| 30 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.nds.vd4dotsu.nxv4i32.nxv16i8.nxv16i8.i64(<vscale x 4 x i32> [[VD:%.*]], <vscale x 16 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 31 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i32> [[TMP0]] |
| 32 |
// |
| 33 |
vint32m2_t test_nds_vd4dotsu_vv_i32m2_tu(vint32m2_t vd, vint8m2_t vs1, vuint8m2_t vs2, size_t vl) { |
| 34 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 35 |
} |
| 36 |
|
| 37 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m4_tu( |
| 38 |
// CHECK-RV64-NEXT: entry: |
| 39 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.nds.vd4dotsu.nxv8i32.nxv32i8.nxv32i8.i64(<vscale x 8 x i32> [[VD:%.*]], <vscale x 32 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 40 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i32> [[TMP0]] |
| 41 |
// |
| 42 |
vint32m4_t test_nds_vd4dotsu_vv_i32m4_tu(vint32m4_t vd, vint8m4_t vs1, vuint8m4_t vs2, size_t vl) { |
| 43 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 44 |
} |
| 45 |
|
| 46 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m8_tu( |
| 47 |
// CHECK-RV64-NEXT: entry: |
| 48 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.nds.vd4dotsu.nxv16i32.nxv64i8.nxv64i8.i64(<vscale x 16 x i32> [[VD:%.*]], <vscale x 64 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 49 |
// CHECK-RV64-NEXT: ret <vscale x 16 x i32> [[TMP0]] |
| 50 |
// |
| 51 |
vint32m8_t test_nds_vd4dotsu_vv_i32m8_tu(vint32m8_t vd, vint8m8_t vs1, vuint8m8_t vs2, size_t vl) { |
| 52 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 53 |
} |
| 54 |
|
| 55 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m1_tu( |
| 56 |
// CHECK-RV64-NEXT: entry: |
| 57 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.nds.vd4dotsu.nxv1i64.nxv4i16.nxv4i16.i64(<vscale x 1 x i64> [[VD:%.*]], <vscale x 4 x i16> [[VS1:%.*]], <vscale x 4 x i16> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 58 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i64> [[TMP0]] |
| 59 |
// |
| 60 |
vint64m1_t test_nds_vd4dotsu_vv_i64m1_tu(vint64m1_t vd, vint16m1_t vs1, vuint16m1_t vs2, size_t vl) { |
| 61 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 62 |
} |
| 63 |
|
| 64 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m2_tu( |
| 65 |
// CHECK-RV64-NEXT: entry: |
| 66 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.nds.vd4dotsu.nxv2i64.nxv8i16.nxv8i16.i64(<vscale x 2 x i64> [[VD:%.*]], <vscale x 8 x i16> [[VS1:%.*]], <vscale x 8 x i16> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 67 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i64> [[TMP0]] |
| 68 |
// |
| 69 |
vint64m2_t test_nds_vd4dotsu_vv_i64m2_tu(vint64m2_t vd, vint16m2_t vs1, vuint16m2_t vs2, size_t vl) { |
| 70 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 71 |
} |
| 72 |
|
| 73 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m4_tu( |
| 74 |
// CHECK-RV64-NEXT: entry: |
| 75 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.nds.vd4dotsu.nxv4i64.nxv16i16.nxv16i16.i64(<vscale x 4 x i64> [[VD:%.*]], <vscale x 16 x i16> [[VS1:%.*]], <vscale x 16 x i16> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 76 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i64> [[TMP0]] |
| 77 |
// |
| 78 |
vint64m4_t test_nds_vd4dotsu_vv_i64m4_tu(vint64m4_t vd, vint16m4_t vs1, vuint16m4_t vs2, size_t vl) { |
| 79 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 80 |
} |
| 81 |
|
| 82 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m8_tu( |
| 83 |
// CHECK-RV64-NEXT: entry: |
| 84 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.nds.vd4dotsu.nxv8i64.nxv32i16.nxv32i16.i64(<vscale x 8 x i64> [[VD:%.*]], <vscale x 32 x i16> [[VS1:%.*]], <vscale x 32 x i16> [[VS2:%.*]], i64 [[VL:%.*]], i64 2) |
| 85 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i64> [[TMP0]] |
| 86 |
// |
| 87 |
vint64m8_t test_nds_vd4dotsu_vv_i64m8_tu(vint64m8_t vd, vint16m8_t vs1, vuint16m8_t vs2, size_t vl) { |
| 88 |
return __riscv_nds_vd4dotsu_tu(vd, vs1, vs2, vl); |
| 89 |
} |
| 90 |
|
| 91 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32mf2_tum( |
| 92 |
// CHECK-RV64-NEXT: entry: |
| 93 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv1i32.nxv4i8.nxv4i8.i64(<vscale x 1 x i32> [[VD:%.*]], <vscale x 4 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 94 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i32> [[TMP0]] |
| 95 |
// |
| 96 |
vint32mf2_t test_nds_vd4dotsu_vv_i32mf2_tum(vbool64_t mask, vint32mf2_t vd, vint8mf2_t vs1, vuint8mf2_t vs2, size_t vl) { |
| 97 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 98 |
} |
| 99 |
|
| 100 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m1_tum( |
| 101 |
// CHECK-RV64-NEXT: entry: |
| 102 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 103 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i32> [[TMP0]] |
| 104 |
// |
| 105 |
vint32m1_t test_nds_vd4dotsu_vv_i32m1_tum(vbool32_t mask, vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) { |
| 106 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 107 |
} |
| 108 |
|
| 109 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m2_tum( |
| 110 |
// CHECK-RV64-NEXT: entry: |
| 111 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv4i32.nxv16i8.nxv16i8.i64(<vscale x 4 x i32> [[VD:%.*]], <vscale x 16 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 112 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i32> [[TMP0]] |
| 113 |
// |
| 114 |
vint32m2_t test_nds_vd4dotsu_vv_i32m2_tum(vbool16_t mask, vint32m2_t vd, vint8m2_t vs1, vuint8m2_t vs2, size_t vl) { |
| 115 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 116 |
} |
| 117 |
|
| 118 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m4_tum( |
| 119 |
// CHECK-RV64-NEXT: entry: |
| 120 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv8i32.nxv32i8.nxv32i8.i64(<vscale x 8 x i32> [[VD:%.*]], <vscale x 32 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 121 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i32> [[TMP0]] |
| 122 |
// |
| 123 |
vint32m4_t test_nds_vd4dotsu_vv_i32m4_tum(vbool8_t mask, vint32m4_t vd, vint8m4_t vs1, vuint8m4_t vs2, size_t vl) { |
| 124 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 125 |
} |
| 126 |
|
| 127 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m8_tum( |
| 128 |
// CHECK-RV64-NEXT: entry: |
| 129 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv16i32.nxv64i8.nxv64i8.i64(<vscale x 16 x i32> [[VD:%.*]], <vscale x 64 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 130 |
// CHECK-RV64-NEXT: ret <vscale x 16 x i32> [[TMP0]] |
| 131 |
// |
| 132 |
vint32m8_t test_nds_vd4dotsu_vv_i32m8_tum(vbool4_t mask, vint32m8_t vd, vint8m8_t vs1, vuint8m8_t vs2, size_t vl) { |
| 133 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 134 |
} |
| 135 |
|
| 136 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m1_tum( |
| 137 |
// CHECK-RV64-NEXT: entry: |
| 138 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv1i64.nxv4i16.nxv4i16.i64(<vscale x 1 x i64> [[VD:%.*]], <vscale x 4 x i16> [[VS1:%.*]], <vscale x 4 x i16> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 139 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i64> [[TMP0]] |
| 140 |
// |
| 141 |
vint64m1_t test_nds_vd4dotsu_vv_i64m1_tum(vbool64_t mask, vint64m1_t vd, vint16m1_t vs1, vuint16m1_t vs2, size_t vl) { |
| 142 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 143 |
} |
| 144 |
|
| 145 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m2_tum( |
| 146 |
// CHECK-RV64-NEXT: entry: |
| 147 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv2i64.nxv8i16.nxv8i16.i64(<vscale x 2 x i64> [[VD:%.*]], <vscale x 8 x i16> [[VS1:%.*]], <vscale x 8 x i16> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 148 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i64> [[TMP0]] |
| 149 |
// |
| 150 |
vint64m2_t test_nds_vd4dotsu_vv_i64m2_tum(vbool32_t mask, vint64m2_t vd, vint16m2_t vs1, vuint16m2_t vs2, size_t vl) { |
| 151 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 152 |
} |
| 153 |
|
| 154 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m4_tum( |
| 155 |
// CHECK-RV64-NEXT: entry: |
| 156 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv4i64.nxv16i16.nxv16i16.i64(<vscale x 4 x i64> [[VD:%.*]], <vscale x 16 x i16> [[VS1:%.*]], <vscale x 16 x i16> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 157 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i64> [[TMP0]] |
| 158 |
// |
| 159 |
vint64m4_t test_nds_vd4dotsu_vv_i64m4_tum(vbool16_t mask, vint64m4_t vd, vint16m4_t vs1, vuint16m4_t vs2, size_t vl) { |
| 160 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 161 |
} |
| 162 |
|
| 163 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m8_tum( |
| 164 |
// CHECK-RV64-NEXT: entry: |
| 165 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv8i64.nxv32i16.nxv32i16.i64(<vscale x 8 x i64> [[VD:%.*]], <vscale x 32 x i16> [[VS1:%.*]], <vscale x 32 x i16> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 2) |
| 166 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i64> [[TMP0]] |
| 167 |
// |
| 168 |
vint64m8_t test_nds_vd4dotsu_vv_i64m8_tum(vbool8_t mask, vint64m8_t vd, vint16m8_t vs1, vuint16m8_t vs2, size_t vl) { |
| 169 |
return __riscv_nds_vd4dotsu_tum(mask, vd, vs1, vs2, vl); |
| 170 |
} |
| 171 |
|
| 172 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32mf2_tumu( |
| 173 |
// CHECK-RV64-NEXT: entry: |
| 174 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv1i32.nxv4i8.nxv4i8.i64(<vscale x 1 x i32> [[VD:%.*]], <vscale x 4 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 175 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i32> [[TMP0]] |
| 176 |
// |
| 177 |
vint32mf2_t test_nds_vd4dotsu_vv_i32mf2_tumu(vbool64_t mask, vint32mf2_t vd, vint8mf2_t vs1, vuint8mf2_t vs2, size_t vl) { |
| 178 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 179 |
} |
| 180 |
|
| 181 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m1_tumu( |
| 182 |
// CHECK-RV64-NEXT: entry: |
| 183 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 184 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i32> [[TMP0]] |
| 185 |
// |
| 186 |
vint32m1_t test_nds_vd4dotsu_vv_i32m1_tumu(vbool32_t mask, vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) { |
| 187 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 188 |
} |
| 189 |
|
| 190 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m2_tumu( |
| 191 |
// CHECK-RV64-NEXT: entry: |
| 192 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv4i32.nxv16i8.nxv16i8.i64(<vscale x 4 x i32> [[VD:%.*]], <vscale x 16 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 193 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i32> [[TMP0]] |
| 194 |
// |
| 195 |
vint32m2_t test_nds_vd4dotsu_vv_i32m2_tumu(vbool16_t mask, vint32m2_t vd, vint8m2_t vs1, vuint8m2_t vs2, size_t vl) { |
| 196 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 197 |
} |
| 198 |
|
| 199 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m4_tumu( |
| 200 |
// CHECK-RV64-NEXT: entry: |
| 201 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv8i32.nxv32i8.nxv32i8.i64(<vscale x 8 x i32> [[VD:%.*]], <vscale x 32 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 202 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i32> [[TMP0]] |
| 203 |
// |
| 204 |
vint32m4_t test_nds_vd4dotsu_vv_i32m4_tumu(vbool8_t mask, vint32m4_t vd, vint8m4_t vs1, vuint8m4_t vs2, size_t vl) { |
| 205 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 206 |
} |
| 207 |
|
| 208 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m8_tumu( |
| 209 |
// CHECK-RV64-NEXT: entry: |
| 210 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv16i32.nxv64i8.nxv64i8.i64(<vscale x 16 x i32> [[VD:%.*]], <vscale x 64 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 211 |
// CHECK-RV64-NEXT: ret <vscale x 16 x i32> [[TMP0]] |
| 212 |
// |
| 213 |
vint32m8_t test_nds_vd4dotsu_vv_i32m8_tumu(vbool4_t mask, vint32m8_t vd, vint8m8_t vs1, vuint8m8_t vs2, size_t vl) { |
| 214 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 215 |
} |
| 216 |
|
| 217 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m1_tumu( |
| 218 |
// CHECK-RV64-NEXT: entry: |
| 219 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv1i64.nxv4i16.nxv4i16.i64(<vscale x 1 x i64> [[VD:%.*]], <vscale x 4 x i16> [[VS1:%.*]], <vscale x 4 x i16> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 220 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i64> [[TMP0]] |
| 221 |
// |
| 222 |
vint64m1_t test_nds_vd4dotsu_vv_i64m1_tumu(vbool64_t mask, vint64m1_t vd, vint16m1_t vs1, vuint16m1_t vs2, size_t vl) { |
| 223 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 224 |
} |
| 225 |
|
| 226 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m2_tumu( |
| 227 |
// CHECK-RV64-NEXT: entry: |
| 228 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv2i64.nxv8i16.nxv8i16.i64(<vscale x 2 x i64> [[VD:%.*]], <vscale x 8 x i16> [[VS1:%.*]], <vscale x 8 x i16> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 229 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i64> [[TMP0]] |
| 230 |
// |
| 231 |
vint64m2_t test_nds_vd4dotsu_vv_i64m2_tumu(vbool32_t mask, vint64m2_t vd, vint16m2_t vs1, vuint16m2_t vs2, size_t vl) { |
| 232 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 233 |
} |
| 234 |
|
| 235 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m4_tumu( |
| 236 |
// CHECK-RV64-NEXT: entry: |
| 237 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv4i64.nxv16i16.nxv16i16.i64(<vscale x 4 x i64> [[VD:%.*]], <vscale x 16 x i16> [[VS1:%.*]], <vscale x 16 x i16> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 238 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i64> [[TMP0]] |
| 239 |
// |
| 240 |
vint64m4_t test_nds_vd4dotsu_vv_i64m4_tumu(vbool16_t mask, vint64m4_t vd, vint16m4_t vs1, vuint16m4_t vs2, size_t vl) { |
| 241 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 242 |
} |
| 243 |
|
| 244 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m8_tumu( |
| 245 |
// CHECK-RV64-NEXT: entry: |
| 246 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv8i64.nxv32i16.nxv32i16.i64(<vscale x 8 x i64> [[VD:%.*]], <vscale x 32 x i16> [[VS1:%.*]], <vscale x 32 x i16> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0) |
| 247 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i64> [[TMP0]] |
| 248 |
// |
| 249 |
vint64m8_t test_nds_vd4dotsu_vv_i64m8_tumu(vbool8_t mask, vint64m8_t vd, vint16m8_t vs1, vuint16m8_t vs2, size_t vl) { |
| 250 |
return __riscv_nds_vd4dotsu_tumu(mask, vd, vs1, vs2, vl); |
| 251 |
} |
| 252 |
|
| 253 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32mf2_mu( |
| 254 |
// CHECK-RV64-NEXT: entry: |
| 255 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv1i32.nxv4i8.nxv4i8.i64(<vscale x 1 x i32> [[VD:%.*]], <vscale x 4 x i8> [[VS1:%.*]], <vscale x 4 x i8> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 256 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i32> [[TMP0]] |
| 257 |
// |
| 258 |
vint32mf2_t test_nds_vd4dotsu_vv_i32mf2_mu(vbool64_t mask, vint32mf2_t vd, vint8mf2_t vs1, vuint8mf2_t vs2, size_t vl) { |
| 259 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 260 |
} |
| 261 |
|
| 262 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m1_mu( |
| 263 |
// CHECK-RV64-NEXT: entry: |
| 264 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv2i32.nxv8i8.nxv8i8.i64(<vscale x 2 x i32> [[VD:%.*]], <vscale x 8 x i8> [[VS1:%.*]], <vscale x 8 x i8> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 265 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i32> [[TMP0]] |
| 266 |
// |
| 267 |
vint32m1_t test_nds_vd4dotsu_vv_i32m1_mu(vbool32_t mask, vint32m1_t vd, vint8m1_t vs1, vuint8m1_t vs2, size_t vl) { |
| 268 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 269 |
} |
| 270 |
|
| 271 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m2_mu( |
| 272 |
// CHECK-RV64-NEXT: entry: |
| 273 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv4i32.nxv16i8.nxv16i8.i64(<vscale x 4 x i32> [[VD:%.*]], <vscale x 16 x i8> [[VS1:%.*]], <vscale x 16 x i8> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 274 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i32> [[TMP0]] |
| 275 |
// |
| 276 |
vint32m2_t test_nds_vd4dotsu_vv_i32m2_mu(vbool16_t mask, vint32m2_t vd, vint8m2_t vs1, vuint8m2_t vs2, size_t vl) { |
| 277 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 278 |
} |
| 279 |
|
| 280 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m4_mu( |
| 281 |
// CHECK-RV64-NEXT: entry: |
| 282 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv8i32.nxv32i8.nxv32i8.i64(<vscale x 8 x i32> [[VD:%.*]], <vscale x 32 x i8> [[VS1:%.*]], <vscale x 32 x i8> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 283 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i32> [[TMP0]] |
| 284 |
// |
| 285 |
vint32m4_t test_nds_vd4dotsu_vv_i32m4_mu(vbool8_t mask, vint32m4_t vd, vint8m4_t vs1, vuint8m4_t vs2, size_t vl) { |
| 286 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 287 |
} |
| 288 |
|
| 289 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i32m8_mu( |
| 290 |
// CHECK-RV64-NEXT: entry: |
| 291 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 16 x i32> @llvm.riscv.nds.vd4dotsu.mask.nxv16i32.nxv64i8.nxv64i8.i64(<vscale x 16 x i32> [[VD:%.*]], <vscale x 64 x i8> [[VS1:%.*]], <vscale x 64 x i8> [[VS2:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 292 |
// CHECK-RV64-NEXT: ret <vscale x 16 x i32> [[TMP0]] |
| 293 |
// |
| 294 |
vint32m8_t test_nds_vd4dotsu_vv_i32m8_mu(vbool4_t mask, vint32m8_t vd, vint8m8_t vs1, vuint8m8_t vs2, size_t vl) { |
| 295 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 296 |
} |
| 297 |
|
| 298 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m1_mu( |
| 299 |
// CHECK-RV64-NEXT: entry: |
| 300 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 1 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv1i64.nxv4i16.nxv4i16.i64(<vscale x 1 x i64> [[VD:%.*]], <vscale x 4 x i16> [[VS1:%.*]], <vscale x 4 x i16> [[VS2:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 301 |
// CHECK-RV64-NEXT: ret <vscale x 1 x i64> [[TMP0]] |
| 302 |
// |
| 303 |
vint64m1_t test_nds_vd4dotsu_vv_i64m1_mu(vbool64_t mask, vint64m1_t vd, vint16m1_t vs1, vuint16m1_t vs2, size_t vl) { |
| 304 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 305 |
} |
| 306 |
|
| 307 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m2_mu( |
| 308 |
// CHECK-RV64-NEXT: entry: |
| 309 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 2 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv2i64.nxv8i16.nxv8i16.i64(<vscale x 2 x i64> [[VD:%.*]], <vscale x 8 x i16> [[VS1:%.*]], <vscale x 8 x i16> [[VS2:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 310 |
// CHECK-RV64-NEXT: ret <vscale x 2 x i64> [[TMP0]] |
| 311 |
// |
| 312 |
vint64m2_t test_nds_vd4dotsu_vv_i64m2_mu(vbool32_t mask, vint64m2_t vd, vint16m2_t vs1, vuint16m2_t vs2, size_t vl) { |
| 313 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 314 |
} |
| 315 |
|
| 316 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m4_mu( |
| 317 |
// CHECK-RV64-NEXT: entry: |
| 318 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 4 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv4i64.nxv16i16.nxv16i16.i64(<vscale x 4 x i64> [[VD:%.*]], <vscale x 16 x i16> [[VS1:%.*]], <vscale x 16 x i16> [[VS2:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 319 |
// CHECK-RV64-NEXT: ret <vscale x 4 x i64> [[TMP0]] |
| 320 |
// |
| 321 |
vint64m4_t test_nds_vd4dotsu_vv_i64m4_mu(vbool16_t mask, vint64m4_t vd, vint16m4_t vs1, vuint16m4_t vs2, size_t vl) { |
| 322 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 323 |
} |
| 324 |
|
| 325 |
// CHECK-RV64-LABEL: @test_nds_vd4dotsu_vv_i64m8_mu( |
| 326 |
// CHECK-RV64-NEXT: entry: |
| 327 |
// CHECK-RV64-NEXT: [[TMP0:%.*]] = call <vscale x 8 x i64> @llvm.riscv.nds.vd4dotsu.mask.nxv8i64.nxv32i16.nxv32i16.i64(<vscale x 8 x i64> [[VD:%.*]], <vscale x 32 x i16> [[VS1:%.*]], <vscale x 32 x i16> [[VS2:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 1) |
| 328 |
// CHECK-RV64-NEXT: ret <vscale x 8 x i64> [[TMP0]] |
| 329 |
// |
| 330 |
vint64m8_t test_nds_vd4dotsu_vv_i64m8_mu(vbool8_t mask, vint64m8_t vd, vint16m8_t vs1, vuint16m8_t vs2, size_t vl) { |
| 331 |
return __riscv_nds_vd4dotsu_mu(mask, vd, vs1, vs2, vl); |
| 332 |
} |