1 // RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \ 2 // RUN: -ffp-contract=fast -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg \ 3 // RUN: | FileCheck %s 4 5 // Test new aarch64 intrinsics with poly64 6 7 #include <arm_neon.h> 8 9 // CHECK-LABEL: define <1 x i64> @test_vceq_p64(<1 x i64> %a, <1 x i64> %b) #0 { 10 // CHECK: [[CMP_I:%.*]] = icmp eq <1 x i64> %a, %b 11 // CHECK: [[SEXT_I:%.*]] = sext <1 x i1> [[CMP_I]] to <1 x i64> 12 // CHECK: ret <1 x i64> [[SEXT_I]] 13 uint64x1_t test_vceq_p64(poly64x1_t a, poly64x1_t b) { 14 return vceq_p64(a, b); 15 } 16 17 // CHECK-LABEL: define <2 x i64> @test_vceqq_p64(<2 x i64> %a, <2 x i64> %b) #1 { 18 // CHECK: [[CMP_I:%.*]] = icmp eq <2 x i64> %a, %b 19 // CHECK: [[SEXT_I:%.*]] = sext <2 x i1> [[CMP_I]] to <2 x i64> 20 // CHECK: ret <2 x i64> [[SEXT_I]] 21 uint64x2_t test_vceqq_p64(poly64x2_t a, poly64x2_t b) { 22 return vceqq_p64(a, b); 23 } 24 25 // CHECK-LABEL: define <1 x i64> @test_vtst_p64(<1 x i64> %a, <1 x i64> %b) #0 { 26 // CHECK: [[TMP4:%.*]] = and <1 x i64> %a, %b 27 // CHECK: [[TMP5:%.*]] = icmp ne <1 x i64> [[TMP4]], zeroinitializer 28 // CHECK: [[VTST_I:%.*]] = sext <1 x i1> [[TMP5]] to <1 x i64> 29 // CHECK: ret <1 x i64> [[VTST_I]] 30 uint64x1_t test_vtst_p64(poly64x1_t a, poly64x1_t b) { 31 return vtst_p64(a, b); 32 } 33 34 // CHECK-LABEL: define <2 x i64> @test_vtstq_p64(<2 x i64> %a, <2 x i64> %b) #1 { 35 // CHECK: [[TMP4:%.*]] = and <2 x i64> %a, %b 36 // CHECK: [[TMP5:%.*]] = icmp ne <2 x i64> [[TMP4]], zeroinitializer 37 // CHECK: [[VTST_I:%.*]] = sext <2 x i1> [[TMP5]] to <2 x i64> 38 // CHECK: ret <2 x i64> [[VTST_I]] 39 uint64x2_t test_vtstq_p64(poly64x2_t a, poly64x2_t b) { 40 return vtstq_p64(a, b); 41 } 42 43 // CHECK-LABEL: define <1 x i64> @test_vbsl_p64(<1 x i64> %a, <1 x i64> %b, <1 x i64> %c) #0 { 44 // CHECK: [[VBSL3_I:%.*]] = and <1 x i64> %a, %b 45 // CHECK: [[TMP3:%.*]] = xor <1 x i64> %a, <i64 -1> 46 // CHECK: [[VBSL4_I:%.*]] = and <1 x i64> [[TMP3]], %c 47 // CHECK: [[VBSL5_I:%.*]] = or <1 x i64> [[VBSL3_I]], [[VBSL4_I]] 48 // CHECK: ret <1 x i64> [[VBSL5_I]] 49 poly64x1_t test_vbsl_p64(poly64x1_t a, poly64x1_t b, poly64x1_t c) { 50 return vbsl_p64(a, b, c); 51 } 52 53 // CHECK-LABEL: define <2 x i64> @test_vbslq_p64(<2 x i64> %a, <2 x i64> %b, <2 x i64> %c) #1 { 54 // CHECK: [[VBSL3_I:%.*]] = and <2 x i64> %a, %b 55 // CHECK: [[TMP3:%.*]] = xor <2 x i64> %a, <i64 -1, i64 -1> 56 // CHECK: [[VBSL4_I:%.*]] = and <2 x i64> [[TMP3]], %c 57 // CHECK: [[VBSL5_I:%.*]] = or <2 x i64> [[VBSL3_I]], [[VBSL4_I]] 58 // CHECK: ret <2 x i64> [[VBSL5_I]] 59 poly64x2_t test_vbslq_p64(poly64x2_t a, poly64x2_t b, poly64x2_t c) { 60 return vbslq_p64(a, b, c); 61 } 62 63 // CHECK-LABEL: define i64 @test_vget_lane_p64(<1 x i64> %v) #0 { 64 // CHECK: [[VGET_LANE:%.*]] = extractelement <1 x i64> %v, i32 0 65 // CHECK: ret i64 [[VGET_LANE]] 66 poly64_t test_vget_lane_p64(poly64x1_t v) { 67 return vget_lane_p64(v, 0); 68 } 69 70 // CHECK-LABEL: define i64 @test_vgetq_lane_p64(<2 x i64> %v) #1 { 71 // CHECK: [[VGETQ_LANE:%.*]] = extractelement <2 x i64> %v, i32 1 72 // CHECK: ret i64 [[VGETQ_LANE]] 73 poly64_t test_vgetq_lane_p64(poly64x2_t v) { 74 return vgetq_lane_p64(v, 1); 75 } 76 77 // CHECK-LABEL: define <1 x i64> @test_vset_lane_p64(i64 %a, <1 x i64> %v) #0 { 78 // CHECK: [[VSET_LANE:%.*]] = insertelement <1 x i64> %v, i64 %a, i32 0 79 // CHECK: ret <1 x i64> [[VSET_LANE]] 80 poly64x1_t test_vset_lane_p64(poly64_t a, poly64x1_t v) { 81 return vset_lane_p64(a, v, 0); 82 } 83 84 // CHECK-LABEL: define <2 x i64> @test_vsetq_lane_p64(i64 %a, <2 x i64> %v) #1 { 85 // CHECK: [[VSET_LANE:%.*]] = insertelement <2 x i64> %v, i64 %a, i32 1 86 // CHECK: ret <2 x i64> [[VSET_LANE]] 87 poly64x2_t test_vsetq_lane_p64(poly64_t a, poly64x2_t v) { 88 return vsetq_lane_p64(a, v, 1); 89 } 90 91 // CHECK-LABEL: define <1 x i64> @test_vcopy_lane_p64(<1 x i64> %a, <1 x i64> %b) #0 { 92 // CHECK: [[VGET_LANE:%.*]] = extractelement <1 x i64> %b, i32 0 93 // CHECK: [[VSET_LANE:%.*]] = insertelement <1 x i64> %a, i64 [[VGET_LANE]], i32 0 94 // CHECK: ret <1 x i64> [[VSET_LANE]] 95 poly64x1_t test_vcopy_lane_p64(poly64x1_t a, poly64x1_t b) { 96 return vcopy_lane_p64(a, 0, b, 0); 97 98 } 99 100 // CHECK-LABEL: define <2 x i64> @test_vcopyq_lane_p64(<2 x i64> %a, <1 x i64> %b) #1 { 101 // CHECK: [[VGET_LANE:%.*]] = extractelement <1 x i64> %b, i32 0 102 // CHECK: [[VSET_LANE:%.*]] = insertelement <2 x i64> %a, i64 [[VGET_LANE]], i32 1 103 // CHECK: ret <2 x i64> [[VSET_LANE]] 104 poly64x2_t test_vcopyq_lane_p64(poly64x2_t a, poly64x1_t b) { 105 return vcopyq_lane_p64(a, 1, b, 0); 106 } 107 108 // CHECK-LABEL: define <2 x i64> @test_vcopyq_laneq_p64(<2 x i64> %a, <2 x i64> %b) #1 { 109 // CHECK: [[VGETQ_LANE:%.*]] = extractelement <2 x i64> %b, i32 1 110 // CHECK: [[VSET_LANE:%.*]] = insertelement <2 x i64> %a, i64 [[VGETQ_LANE]], i32 1 111 // CHECK: ret <2 x i64> [[VSET_LANE]] 112 poly64x2_t test_vcopyq_laneq_p64(poly64x2_t a, poly64x2_t b) { 113 return vcopyq_laneq_p64(a, 1, b, 1); 114 } 115 116 // CHECK-LABEL: define <1 x i64> @test_vcreate_p64(i64 %a) #0 { 117 // CHECK: [[TMP0:%.*]] = bitcast i64 %a to <1 x i64> 118 // CHECK: ret <1 x i64> [[TMP0]] 119 poly64x1_t test_vcreate_p64(uint64_t a) { 120 return vcreate_p64(a); 121 } 122 123 // CHECK-LABEL: define <1 x i64> @test_vdup_n_p64(i64 %a) #0 { 124 // CHECK: [[VECINIT_I:%.*]] = insertelement <1 x i64> undef, i64 %a, i32 0 125 // CHECK: ret <1 x i64> [[VECINIT_I]] 126 poly64x1_t test_vdup_n_p64(poly64_t a) { 127 return vdup_n_p64(a); 128 } 129 // CHECK-LABEL: define <2 x i64> @test_vdupq_n_p64(i64 %a) #1 { 130 // CHECK: [[VECINIT_I:%.*]] = insertelement <2 x i64> undef, i64 %a, i32 0 131 // CHECK: [[VECINIT1_I:%.*]] = insertelement <2 x i64> [[VECINIT_I]], i64 %a, i32 1 132 // CHECK: ret <2 x i64> [[VECINIT1_I]] 133 poly64x2_t test_vdupq_n_p64(poly64_t a) { 134 return vdupq_n_p64(a); 135 } 136 137 // CHECK-LABEL: define <1 x i64> @test_vmov_n_p64(i64 %a) #0 { 138 // CHECK: [[VECINIT_I:%.*]] = insertelement <1 x i64> undef, i64 %a, i32 0 139 // CHECK: ret <1 x i64> [[VECINIT_I]] 140 poly64x1_t test_vmov_n_p64(poly64_t a) { 141 return vmov_n_p64(a); 142 } 143 144 // CHECK-LABEL: define <2 x i64> @test_vmovq_n_p64(i64 %a) #1 { 145 // CHECK: [[VECINIT_I:%.*]] = insertelement <2 x i64> undef, i64 %a, i32 0 146 // CHECK: [[VECINIT1_I:%.*]] = insertelement <2 x i64> [[VECINIT_I]], i64 %a, i32 1 147 // CHECK: ret <2 x i64> [[VECINIT1_I]] 148 poly64x2_t test_vmovq_n_p64(poly64_t a) { 149 return vmovq_n_p64(a); 150 } 151 152 // CHECK-LABEL: define <1 x i64> @test_vdup_lane_p64(<1 x i64> %vec) #0 { 153 // CHECK: [[SHUFFLE:%.*]] = shufflevector <1 x i64> %vec, <1 x i64> %vec, <1 x i32> zeroinitializer 154 // CHECK: ret <1 x i64> [[SHUFFLE]] 155 poly64x1_t test_vdup_lane_p64(poly64x1_t vec) { 156 return vdup_lane_p64(vec, 0); 157 } 158 159 // CHECK-LABEL: define <2 x i64> @test_vdupq_lane_p64(<1 x i64> %vec) #1 { 160 // CHECK: [[SHUFFLE:%.*]] = shufflevector <1 x i64> %vec, <1 x i64> %vec, <2 x i32> zeroinitializer 161 // CHECK: ret <2 x i64> [[SHUFFLE]] 162 poly64x2_t test_vdupq_lane_p64(poly64x1_t vec) { 163 return vdupq_lane_p64(vec, 0); 164 } 165 166 // CHECK-LABEL: define <2 x i64> @test_vdupq_laneq_p64(<2 x i64> %vec) #1 { 167 // CHECK: [[SHUFFLE:%.*]] = shufflevector <2 x i64> %vec, <2 x i64> %vec, <2 x i32> <i32 1, i32 1> 168 // CHECK: ret <2 x i64> [[SHUFFLE]] 169 poly64x2_t test_vdupq_laneq_p64(poly64x2_t vec) { 170 return vdupq_laneq_p64(vec, 1); 171 } 172 173 // CHECK-LABEL: define <2 x i64> @test_vcombine_p64(<1 x i64> %low, <1 x i64> %high) #1 { 174 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <1 x i64> %low, <1 x i64> %high, <2 x i32> <i32 0, i32 1> 175 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 176 poly64x2_t test_vcombine_p64(poly64x1_t low, poly64x1_t high) { 177 return vcombine_p64(low, high); 178 } 179 180 // CHECK-LABEL: define <1 x i64> @test_vld1_p64(i64* %ptr) #0 { 181 // CHECK: [[TMP0:%.*]] = bitcast i64* %ptr to i8* 182 // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <1 x i64>* 183 // CHECK: [[TMP2:%.*]] = load <1 x i64>, <1 x i64>* [[TMP1]] 184 // CHECK: ret <1 x i64> [[TMP2]] 185 poly64x1_t test_vld1_p64(poly64_t const * ptr) { 186 return vld1_p64(ptr); 187 } 188 189 // CHECK-LABEL: define <2 x i64> @test_vld1q_p64(i64* %ptr) #1 { 190 // CHECK: [[TMP0:%.*]] = bitcast i64* %ptr to i8* 191 // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <2 x i64>* 192 // CHECK: [[TMP2:%.*]] = load <2 x i64>, <2 x i64>* [[TMP1]] 193 // CHECK: ret <2 x i64> [[TMP2]] 194 poly64x2_t test_vld1q_p64(poly64_t const * ptr) { 195 return vld1q_p64(ptr); 196 } 197 198 // CHECK-LABEL: define void @test_vst1_p64(i64* %ptr, <1 x i64> %val) #0 { 199 // CHECK: [[TMP0:%.*]] = bitcast i64* %ptr to i8* 200 // CHECK: [[TMP1:%.*]] = bitcast <1 x i64> %val to <8 x i8> 201 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP0]] to <1 x i64>* 202 // CHECK: [[TMP3:%.*]] = bitcast <8 x i8> [[TMP1]] to <1 x i64> 203 // CHECK: store <1 x i64> [[TMP3]], <1 x i64>* [[TMP2]] 204 // CHECK: ret void 205 void test_vst1_p64(poly64_t * ptr, poly64x1_t val) { 206 return vst1_p64(ptr, val); 207 } 208 209 // CHECK-LABEL: define void @test_vst1q_p64(i64* %ptr, <2 x i64> %val) #1 { 210 // CHECK: [[TMP0:%.*]] = bitcast i64* %ptr to i8* 211 // CHECK: [[TMP1:%.*]] = bitcast <2 x i64> %val to <16 x i8> 212 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP0]] to <2 x i64>* 213 // CHECK: [[TMP3:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x i64> 214 // CHECK: store <2 x i64> [[TMP3]], <2 x i64>* [[TMP2]] 215 // CHECK: ret void 216 void test_vst1q_p64(poly64_t * ptr, poly64x2_t val) { 217 return vst1q_p64(ptr, val); 218 } 219 220 // CHECK-LABEL: define %struct.poly64x1x2_t @test_vld2_p64(i64* %ptr) #2 { 221 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x1x2_t, align 8 222 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x1x2_t, align 8 223 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x2_t* [[__RET]] to i8* 224 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 225 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <1 x i64>* 226 // CHECK: [[VLD2:%.*]] = call { <1 x i64>, <1 x i64> } @llvm.aarch64.neon.ld2.v1i64.p0v1i64(<1 x i64>* [[TMP2]]) 227 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64> }* 228 // CHECK: store { <1 x i64>, <1 x i64> } [[VLD2]], { <1 x i64>, <1 x i64> }* [[TMP3]] 229 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x1x2_t* [[RETVAL]] to i8* 230 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x1x2_t* [[__RET]] to i8* 231 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], i64 16, i1 false) 232 // CHECK: [[TMP6:%.*]] = load %struct.poly64x1x2_t, %struct.poly64x1x2_t* [[RETVAL]], align 8 233 // CHECK: ret %struct.poly64x1x2_t [[TMP6]] 234 poly64x1x2_t test_vld2_p64(poly64_t const * ptr) { 235 return vld2_p64(ptr); 236 } 237 238 // CHECK-LABEL: define %struct.poly64x2x2_t @test_vld2q_p64(i64* %ptr) #2 { 239 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x2x2_t, align 16 240 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x2x2_t, align 16 241 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x2_t* [[__RET]] to i8* 242 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 243 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <2 x i64>* 244 // CHECK: [[VLD2:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.aarch64.neon.ld2.v2i64.p0v2i64(<2 x i64>* [[TMP2]]) 245 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64> }* 246 // CHECK: store { <2 x i64>, <2 x i64> } [[VLD2]], { <2 x i64>, <2 x i64> }* [[TMP3]] 247 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x2x2_t* [[RETVAL]] to i8* 248 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x2x2_t* [[__RET]] to i8* 249 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP4]], i8* align 16 [[TMP5]], i64 32, i1 false) 250 // CHECK: [[TMP6:%.*]] = load %struct.poly64x2x2_t, %struct.poly64x2x2_t* [[RETVAL]], align 16 251 // CHECK: ret %struct.poly64x2x2_t [[TMP6]] 252 poly64x2x2_t test_vld2q_p64(poly64_t const * ptr) { 253 return vld2q_p64(ptr); 254 } 255 256 // CHECK-LABEL: define %struct.poly64x1x3_t @test_vld3_p64(i64* %ptr) #2 { 257 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x1x3_t, align 8 258 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x1x3_t, align 8 259 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x3_t* [[__RET]] to i8* 260 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 261 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <1 x i64>* 262 // CHECK: [[VLD3:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64> } @llvm.aarch64.neon.ld3.v1i64.p0v1i64(<1 x i64>* [[TMP2]]) 263 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64> }* 264 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64> } [[VLD3]], { <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] 265 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x1x3_t* [[RETVAL]] to i8* 266 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x1x3_t* [[__RET]] to i8* 267 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], i64 24, i1 false) 268 // CHECK: [[TMP6:%.*]] = load %struct.poly64x1x3_t, %struct.poly64x1x3_t* [[RETVAL]], align 8 269 // CHECK: ret %struct.poly64x1x3_t [[TMP6]] 270 poly64x1x3_t test_vld3_p64(poly64_t const * ptr) { 271 return vld3_p64(ptr); 272 } 273 274 // CHECK-LABEL: define %struct.poly64x2x3_t @test_vld3q_p64(i64* %ptr) #2 { 275 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x2x3_t, align 16 276 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x2x3_t, align 16 277 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x3_t* [[__RET]] to i8* 278 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 279 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <2 x i64>* 280 // CHECK: [[VLD3:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64> } @llvm.aarch64.neon.ld3.v2i64.p0v2i64(<2 x i64>* [[TMP2]]) 281 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64> }* 282 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64> } [[VLD3]], { <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] 283 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x2x3_t* [[RETVAL]] to i8* 284 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x2x3_t* [[__RET]] to i8* 285 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP4]], i8* align 16 [[TMP5]], i64 48, i1 false) 286 // CHECK: [[TMP6:%.*]] = load %struct.poly64x2x3_t, %struct.poly64x2x3_t* [[RETVAL]], align 16 287 // CHECK: ret %struct.poly64x2x3_t [[TMP6]] 288 poly64x2x3_t test_vld3q_p64(poly64_t const * ptr) { 289 return vld3q_p64(ptr); 290 } 291 292 // CHECK-LABEL: define %struct.poly64x1x4_t @test_vld4_p64(i64* %ptr) #2 { 293 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x1x4_t, align 8 294 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x1x4_t, align 8 295 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x4_t* [[__RET]] to i8* 296 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 297 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <1 x i64>* 298 // CHECK: [[VLD4:%.*]] = call { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } @llvm.aarch64.neon.ld4.v1i64.p0v1i64(<1 x i64>* [[TMP2]]) 299 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* 300 // CHECK: store { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> } [[VLD4]], { <1 x i64>, <1 x i64>, <1 x i64>, <1 x i64> }* [[TMP3]] 301 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x1x4_t* [[RETVAL]] to i8* 302 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x1x4_t* [[__RET]] to i8* 303 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP4]], i8* align 8 [[TMP5]], i64 32, i1 false) 304 // CHECK: [[TMP6:%.*]] = load %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[RETVAL]], align 8 305 // CHECK: ret %struct.poly64x1x4_t [[TMP6]] 306 poly64x1x4_t test_vld4_p64(poly64_t const * ptr) { 307 return vld4_p64(ptr); 308 } 309 310 // CHECK-LABEL: define %struct.poly64x2x4_t @test_vld4q_p64(i64* %ptr) #2 { 311 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly64x2x4_t, align 16 312 // CHECK: [[__RET:%.*]] = alloca %struct.poly64x2x4_t, align 16 313 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x4_t* [[__RET]] to i8* 314 // CHECK: [[TMP1:%.*]] = bitcast i64* %ptr to i8* 315 // CHECK: [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <2 x i64>* 316 // CHECK: [[VLD4:%.*]] = call { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } @llvm.aarch64.neon.ld4.v2i64.p0v2i64(<2 x i64>* [[TMP2]]) 317 // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* 318 // CHECK: store { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> } [[VLD4]], { <2 x i64>, <2 x i64>, <2 x i64>, <2 x i64> }* [[TMP3]] 319 // CHECK: [[TMP4:%.*]] = bitcast %struct.poly64x2x4_t* [[RETVAL]] to i8* 320 // CHECK: [[TMP5:%.*]] = bitcast %struct.poly64x2x4_t* [[__RET]] to i8* 321 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP4]], i8* align 16 [[TMP5]], i64 64, i1 false) 322 // CHECK: [[TMP6:%.*]] = load %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[RETVAL]], align 16 323 // CHECK: ret %struct.poly64x2x4_t [[TMP6]] 324 poly64x2x4_t test_vld4q_p64(poly64_t const * ptr) { 325 return vld4q_p64(ptr); 326 } 327 328 // CHECK-LABEL: define void @test_vst2_p64(i64* %ptr, [2 x <1 x i64>] %val.coerce) #2 { 329 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x1x2_t, align 8 330 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x1x2_t, align 8 331 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x1x2_t, %struct.poly64x1x2_t* [[VAL]], i32 0, i32 0 332 // CHECK: store [2 x <1 x i64>] [[VAL]].coerce, [2 x <1 x i64>]* [[COERCE_DIVE]], align 8 333 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x2_t* [[__S1]] to i8* 334 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x1x2_t* [[VAL]] to i8* 335 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP0]], i8* align 8 [[TMP1]], i64 16, i1 false) 336 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 337 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x1x2_t, %struct.poly64x1x2_t* [[__S1]], i32 0, i32 0 338 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [2 x <1 x i64>], [2 x <1 x i64>]* [[VAL1]], i64 0, i64 0 339 // CHECK: [[TMP3:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX]], align 8 340 // CHECK: [[TMP4:%.*]] = bitcast <1 x i64> [[TMP3]] to <8 x i8> 341 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x1x2_t, %struct.poly64x1x2_t* [[__S1]], i32 0, i32 0 342 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [2 x <1 x i64>], [2 x <1 x i64>]* [[VAL2]], i64 0, i64 1 343 // CHECK: [[TMP5:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX3]], align 8 344 // CHECK: [[TMP6:%.*]] = bitcast <1 x i64> [[TMP5]] to <8 x i8> 345 // CHECK: [[TMP7:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x i64> 346 // CHECK: [[TMP8:%.*]] = bitcast <8 x i8> [[TMP6]] to <1 x i64> 347 // CHECK: call void @llvm.aarch64.neon.st2.v1i64.p0i8(<1 x i64> [[TMP7]], <1 x i64> [[TMP8]], i8* [[TMP2]]) 348 // CHECK: ret void 349 void test_vst2_p64(poly64_t * ptr, poly64x1x2_t val) { 350 return vst2_p64(ptr, val); 351 } 352 353 // CHECK-LABEL: define void @test_vst2q_p64(i64* %ptr, [2 x <2 x i64>] %val.coerce) #2 { 354 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x2x2_t, align 16 355 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x2x2_t, align 16 356 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x2x2_t, %struct.poly64x2x2_t* [[VAL]], i32 0, i32 0 357 // CHECK: store [2 x <2 x i64>] [[VAL]].coerce, [2 x <2 x i64>]* [[COERCE_DIVE]], align 16 358 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x2_t* [[__S1]] to i8* 359 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x2x2_t* [[VAL]] to i8* 360 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP0]], i8* align 16 [[TMP1]], i64 32, i1 false) 361 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 362 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x2x2_t, %struct.poly64x2x2_t* [[__S1]], i32 0, i32 0 363 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [2 x <2 x i64>], [2 x <2 x i64>]* [[VAL1]], i64 0, i64 0 364 // CHECK: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX]], align 16 365 // CHECK: [[TMP4:%.*]] = bitcast <2 x i64> [[TMP3]] to <16 x i8> 366 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x2x2_t, %struct.poly64x2x2_t* [[__S1]], i32 0, i32 0 367 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [2 x <2 x i64>], [2 x <2 x i64>]* [[VAL2]], i64 0, i64 1 368 // CHECK: [[TMP5:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX3]], align 16 369 // CHECK: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP5]] to <16 x i8> 370 // CHECK: [[TMP7:%.*]] = bitcast <16 x i8> [[TMP4]] to <2 x i64> 371 // CHECK: [[TMP8:%.*]] = bitcast <16 x i8> [[TMP6]] to <2 x i64> 372 // CHECK: call void @llvm.aarch64.neon.st2.v2i64.p0i8(<2 x i64> [[TMP7]], <2 x i64> [[TMP8]], i8* [[TMP2]]) 373 // CHECK: ret void 374 void test_vst2q_p64(poly64_t * ptr, poly64x2x2_t val) { 375 return vst2q_p64(ptr, val); 376 } 377 378 // CHECK-LABEL: define void @test_vst3_p64(i64* %ptr, [3 x <1 x i64>] %val.coerce) #2 { 379 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x1x3_t, align 8 380 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x1x3_t, align 8 381 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x1x3_t, %struct.poly64x1x3_t* [[VAL]], i32 0, i32 0 382 // CHECK: store [3 x <1 x i64>] [[VAL]].coerce, [3 x <1 x i64>]* [[COERCE_DIVE]], align 8 383 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x3_t* [[__S1]] to i8* 384 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x1x3_t* [[VAL]] to i8* 385 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP0]], i8* align 8 [[TMP1]], i64 24, i1 false) 386 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 387 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x1x3_t, %struct.poly64x1x3_t* [[__S1]], i32 0, i32 0 388 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [3 x <1 x i64>], [3 x <1 x i64>]* [[VAL1]], i64 0, i64 0 389 // CHECK: [[TMP3:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX]], align 8 390 // CHECK: [[TMP4:%.*]] = bitcast <1 x i64> [[TMP3]] to <8 x i8> 391 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x1x3_t, %struct.poly64x1x3_t* [[__S1]], i32 0, i32 0 392 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [3 x <1 x i64>], [3 x <1 x i64>]* [[VAL2]], i64 0, i64 1 393 // CHECK: [[TMP5:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX3]], align 8 394 // CHECK: [[TMP6:%.*]] = bitcast <1 x i64> [[TMP5]] to <8 x i8> 395 // CHECK: [[VAL4:%.*]] = getelementptr inbounds %struct.poly64x1x3_t, %struct.poly64x1x3_t* [[__S1]], i32 0, i32 0 396 // CHECK: [[ARRAYIDX5:%.*]] = getelementptr inbounds [3 x <1 x i64>], [3 x <1 x i64>]* [[VAL4]], i64 0, i64 2 397 // CHECK: [[TMP7:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX5]], align 8 398 // CHECK: [[TMP8:%.*]] = bitcast <1 x i64> [[TMP7]] to <8 x i8> 399 // CHECK: [[TMP9:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x i64> 400 // CHECK: [[TMP10:%.*]] = bitcast <8 x i8> [[TMP6]] to <1 x i64> 401 // CHECK: [[TMP11:%.*]] = bitcast <8 x i8> [[TMP8]] to <1 x i64> 402 // CHECK: call void @llvm.aarch64.neon.st3.v1i64.p0i8(<1 x i64> [[TMP9]], <1 x i64> [[TMP10]], <1 x i64> [[TMP11]], i8* [[TMP2]]) 403 // CHECK: ret void 404 void test_vst3_p64(poly64_t * ptr, poly64x1x3_t val) { 405 return vst3_p64(ptr, val); 406 } 407 408 // CHECK-LABEL: define void @test_vst3q_p64(i64* %ptr, [3 x <2 x i64>] %val.coerce) #2 { 409 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x2x3_t, align 16 410 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x2x3_t, align 16 411 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x2x3_t, %struct.poly64x2x3_t* [[VAL]], i32 0, i32 0 412 // CHECK: store [3 x <2 x i64>] [[VAL]].coerce, [3 x <2 x i64>]* [[COERCE_DIVE]], align 16 413 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x3_t* [[__S1]] to i8* 414 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x2x3_t* [[VAL]] to i8* 415 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP0]], i8* align 16 [[TMP1]], i64 48, i1 false) 416 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 417 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x2x3_t, %struct.poly64x2x3_t* [[__S1]], i32 0, i32 0 418 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [3 x <2 x i64>], [3 x <2 x i64>]* [[VAL1]], i64 0, i64 0 419 // CHECK: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX]], align 16 420 // CHECK: [[TMP4:%.*]] = bitcast <2 x i64> [[TMP3]] to <16 x i8> 421 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x2x3_t, %struct.poly64x2x3_t* [[__S1]], i32 0, i32 0 422 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [3 x <2 x i64>], [3 x <2 x i64>]* [[VAL2]], i64 0, i64 1 423 // CHECK: [[TMP5:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX3]], align 16 424 // CHECK: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP5]] to <16 x i8> 425 // CHECK: [[VAL4:%.*]] = getelementptr inbounds %struct.poly64x2x3_t, %struct.poly64x2x3_t* [[__S1]], i32 0, i32 0 426 // CHECK: [[ARRAYIDX5:%.*]] = getelementptr inbounds [3 x <2 x i64>], [3 x <2 x i64>]* [[VAL4]], i64 0, i64 2 427 // CHECK: [[TMP7:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX5]], align 16 428 // CHECK: [[TMP8:%.*]] = bitcast <2 x i64> [[TMP7]] to <16 x i8> 429 // CHECK: [[TMP9:%.*]] = bitcast <16 x i8> [[TMP4]] to <2 x i64> 430 // CHECK: [[TMP10:%.*]] = bitcast <16 x i8> [[TMP6]] to <2 x i64> 431 // CHECK: [[TMP11:%.*]] = bitcast <16 x i8> [[TMP8]] to <2 x i64> 432 // CHECK: call void @llvm.aarch64.neon.st3.v2i64.p0i8(<2 x i64> [[TMP9]], <2 x i64> [[TMP10]], <2 x i64> [[TMP11]], i8* [[TMP2]]) 433 // CHECK: ret void 434 void test_vst3q_p64(poly64_t * ptr, poly64x2x3_t val) { 435 return vst3q_p64(ptr, val); 436 } 437 438 // CHECK-LABEL: define void @test_vst4_p64(i64* %ptr, [4 x <1 x i64>] %val.coerce) #2 { 439 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x1x4_t, align 8 440 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x1x4_t, align 8 441 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[VAL]], i32 0, i32 0 442 // CHECK: store [4 x <1 x i64>] [[VAL]].coerce, [4 x <1 x i64>]* [[COERCE_DIVE]], align 8 443 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x1x4_t* [[__S1]] to i8* 444 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x1x4_t* [[VAL]] to i8* 445 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[TMP0]], i8* align 8 [[TMP1]], i64 32, i1 false) 446 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 447 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[__S1]], i32 0, i32 0 448 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x <1 x i64>], [4 x <1 x i64>]* [[VAL1]], i64 0, i64 0 449 // CHECK: [[TMP3:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX]], align 8 450 // CHECK: [[TMP4:%.*]] = bitcast <1 x i64> [[TMP3]] to <8 x i8> 451 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[__S1]], i32 0, i32 0 452 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x <1 x i64>], [4 x <1 x i64>]* [[VAL2]], i64 0, i64 1 453 // CHECK: [[TMP5:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX3]], align 8 454 // CHECK: [[TMP6:%.*]] = bitcast <1 x i64> [[TMP5]] to <8 x i8> 455 // CHECK: [[VAL4:%.*]] = getelementptr inbounds %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[__S1]], i32 0, i32 0 456 // CHECK: [[ARRAYIDX5:%.*]] = getelementptr inbounds [4 x <1 x i64>], [4 x <1 x i64>]* [[VAL4]], i64 0, i64 2 457 // CHECK: [[TMP7:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX5]], align 8 458 // CHECK: [[TMP8:%.*]] = bitcast <1 x i64> [[TMP7]] to <8 x i8> 459 // CHECK: [[VAL6:%.*]] = getelementptr inbounds %struct.poly64x1x4_t, %struct.poly64x1x4_t* [[__S1]], i32 0, i32 0 460 // CHECK: [[ARRAYIDX7:%.*]] = getelementptr inbounds [4 x <1 x i64>], [4 x <1 x i64>]* [[VAL6]], i64 0, i64 3 461 // CHECK: [[TMP9:%.*]] = load <1 x i64>, <1 x i64>* [[ARRAYIDX7]], align 8 462 // CHECK: [[TMP10:%.*]] = bitcast <1 x i64> [[TMP9]] to <8 x i8> 463 // CHECK: [[TMP11:%.*]] = bitcast <8 x i8> [[TMP4]] to <1 x i64> 464 // CHECK: [[TMP12:%.*]] = bitcast <8 x i8> [[TMP6]] to <1 x i64> 465 // CHECK: [[TMP13:%.*]] = bitcast <8 x i8> [[TMP8]] to <1 x i64> 466 // CHECK: [[TMP14:%.*]] = bitcast <8 x i8> [[TMP10]] to <1 x i64> 467 // CHECK: call void @llvm.aarch64.neon.st4.v1i64.p0i8(<1 x i64> [[TMP11]], <1 x i64> [[TMP12]], <1 x i64> [[TMP13]], <1 x i64> [[TMP14]], i8* [[TMP2]]) 468 // CHECK: ret void 469 void test_vst4_p64(poly64_t * ptr, poly64x1x4_t val) { 470 return vst4_p64(ptr, val); 471 } 472 473 // CHECK-LABEL: define void @test_vst4q_p64(i64* %ptr, [4 x <2 x i64>] %val.coerce) #2 { 474 // CHECK: [[VAL:%.*]] = alloca %struct.poly64x2x4_t, align 16 475 // CHECK: [[__S1:%.*]] = alloca %struct.poly64x2x4_t, align 16 476 // CHECK: [[COERCE_DIVE:%.*]] = getelementptr inbounds %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[VAL]], i32 0, i32 0 477 // CHECK: store [4 x <2 x i64>] [[VAL]].coerce, [4 x <2 x i64>]* [[COERCE_DIVE]], align 16 478 // CHECK: [[TMP0:%.*]] = bitcast %struct.poly64x2x4_t* [[__S1]] to i8* 479 // CHECK: [[TMP1:%.*]] = bitcast %struct.poly64x2x4_t* [[VAL]] to i8* 480 // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP0]], i8* align 16 [[TMP1]], i64 64, i1 false) 481 // CHECK: [[TMP2:%.*]] = bitcast i64* %ptr to i8* 482 // CHECK: [[VAL1:%.*]] = getelementptr inbounds %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[__S1]], i32 0, i32 0 483 // CHECK: [[ARRAYIDX:%.*]] = getelementptr inbounds [4 x <2 x i64>], [4 x <2 x i64>]* [[VAL1]], i64 0, i64 0 484 // CHECK: [[TMP3:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX]], align 16 485 // CHECK: [[TMP4:%.*]] = bitcast <2 x i64> [[TMP3]] to <16 x i8> 486 // CHECK: [[VAL2:%.*]] = getelementptr inbounds %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[__S1]], i32 0, i32 0 487 // CHECK: [[ARRAYIDX3:%.*]] = getelementptr inbounds [4 x <2 x i64>], [4 x <2 x i64>]* [[VAL2]], i64 0, i64 1 488 // CHECK: [[TMP5:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX3]], align 16 489 // CHECK: [[TMP6:%.*]] = bitcast <2 x i64> [[TMP5]] to <16 x i8> 490 // CHECK: [[VAL4:%.*]] = getelementptr inbounds %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[__S1]], i32 0, i32 0 491 // CHECK: [[ARRAYIDX5:%.*]] = getelementptr inbounds [4 x <2 x i64>], [4 x <2 x i64>]* [[VAL4]], i64 0, i64 2 492 // CHECK: [[TMP7:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX5]], align 16 493 // CHECK: [[TMP8:%.*]] = bitcast <2 x i64> [[TMP7]] to <16 x i8> 494 // CHECK: [[VAL6:%.*]] = getelementptr inbounds %struct.poly64x2x4_t, %struct.poly64x2x4_t* [[__S1]], i32 0, i32 0 495 // CHECK: [[ARRAYIDX7:%.*]] = getelementptr inbounds [4 x <2 x i64>], [4 x <2 x i64>]* [[VAL6]], i64 0, i64 3 496 // CHECK: [[TMP9:%.*]] = load <2 x i64>, <2 x i64>* [[ARRAYIDX7]], align 16 497 // CHECK: [[TMP10:%.*]] = bitcast <2 x i64> [[TMP9]] to <16 x i8> 498 // CHECK: [[TMP11:%.*]] = bitcast <16 x i8> [[TMP4]] to <2 x i64> 499 // CHECK: [[TMP12:%.*]] = bitcast <16 x i8> [[TMP6]] to <2 x i64> 500 // CHECK: [[TMP13:%.*]] = bitcast <16 x i8> [[TMP8]] to <2 x i64> 501 // CHECK: [[TMP14:%.*]] = bitcast <16 x i8> [[TMP10]] to <2 x i64> 502 // CHECK: call void @llvm.aarch64.neon.st4.v2i64.p0i8(<2 x i64> [[TMP11]], <2 x i64> [[TMP12]], <2 x i64> [[TMP13]], <2 x i64> [[TMP14]], i8* [[TMP2]]) 503 // CHECK: ret void 504 void test_vst4q_p64(poly64_t * ptr, poly64x2x4_t val) { 505 return vst4q_p64(ptr, val); 506 } 507 508 // CHECK-LABEL: define <1 x i64> @test_vext_p64(<1 x i64> %a, <1 x i64> %b) #0 { 509 // CHECK: [[TMP0:%.*]] = bitcast <1 x i64> %a to <8 x i8> 510 // CHECK: [[TMP1:%.*]] = bitcast <1 x i64> %b to <8 x i8> 511 // CHECK: [[TMP2:%.*]] = bitcast <8 x i8> [[TMP0]] to <1 x i64> 512 // CHECK: [[TMP3:%.*]] = bitcast <8 x i8> [[TMP1]] to <1 x i64> 513 // CHECK: [[VEXT:%.*]] = shufflevector <1 x i64> [[TMP2]], <1 x i64> [[TMP3]], <1 x i32> zeroinitializer 514 // CHECK: ret <1 x i64> [[VEXT]] 515 poly64x1_t test_vext_p64(poly64x1_t a, poly64x1_t b) { 516 return vext_u64(a, b, 0); 517 518 } 519 520 // CHECK-LABEL: define <2 x i64> @test_vextq_p64(<2 x i64> %a, <2 x i64> %b) #1 { 521 // CHECK: [[TMP0:%.*]] = bitcast <2 x i64> %a to <16 x i8> 522 // CHECK: [[TMP1:%.*]] = bitcast <2 x i64> %b to <16 x i8> 523 // CHECK: [[TMP2:%.*]] = bitcast <16 x i8> [[TMP0]] to <2 x i64> 524 // CHECK: [[TMP3:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x i64> 525 // CHECK: [[VEXT:%.*]] = shufflevector <2 x i64> [[TMP2]], <2 x i64> [[TMP3]], <2 x i32> <i32 1, i32 2> 526 // CHECK: ret <2 x i64> [[VEXT]] 527 poly64x2_t test_vextq_p64(poly64x2_t a, poly64x2_t b) { 528 return vextq_p64(a, b, 1); 529 } 530 531 // CHECK-LABEL: define <2 x i64> @test_vzip1q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 532 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 0, i32 2> 533 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 534 poly64x2_t test_vzip1q_p64(poly64x2_t a, poly64x2_t b) { 535 return vzip1q_p64(a, b); 536 } 537 538 // CHECK-LABEL: define <2 x i64> @test_vzip2q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 539 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 1, i32 3> 540 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 541 poly64x2_t test_vzip2q_p64(poly64x2_t a, poly64x2_t b) { 542 return vzip2q_u64(a, b); 543 } 544 545 // CHECK-LABEL: define <2 x i64> @test_vuzp1q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 546 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 0, i32 2> 547 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 548 poly64x2_t test_vuzp1q_p64(poly64x2_t a, poly64x2_t b) { 549 return vuzp1q_p64(a, b); 550 } 551 552 // CHECK-LABEL: define <2 x i64> @test_vuzp2q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 553 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 1, i32 3> 554 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 555 poly64x2_t test_vuzp2q_p64(poly64x2_t a, poly64x2_t b) { 556 return vuzp2q_u64(a, b); 557 } 558 559 // CHECK-LABEL: define <2 x i64> @test_vtrn1q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 560 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 0, i32 2> 561 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 562 poly64x2_t test_vtrn1q_p64(poly64x2_t a, poly64x2_t b) { 563 return vtrn1q_p64(a, b); 564 } 565 566 // CHECK-LABEL: define <2 x i64> @test_vtrn2q_p64(<2 x i64> %a, <2 x i64> %b) #1 { 567 // CHECK: [[SHUFFLE_I:%.*]] = shufflevector <2 x i64> %a, <2 x i64> %b, <2 x i32> <i32 1, i32 3> 568 // CHECK: ret <2 x i64> [[SHUFFLE_I]] 569 poly64x2_t test_vtrn2q_p64(poly64x2_t a, poly64x2_t b) { 570 return vtrn2q_u64(a, b); 571 } 572 573 // CHECK-LABEL: define <1 x i64> @test_vsri_n_p64(<1 x i64> %a, <1 x i64> %b) #0 { 574 // CHECK: [[TMP0:%.*]] = bitcast <1 x i64> %a to <8 x i8> 575 // CHECK: [[TMP1:%.*]] = bitcast <1 x i64> %b to <8 x i8> 576 // CHECK: [[VSRI_N:%.*]] = bitcast <8 x i8> [[TMP0]] to <1 x i64> 577 // CHECK: [[VSRI_N1:%.*]] = bitcast <8 x i8> [[TMP1]] to <1 x i64> 578 // CHECK: [[VSRI_N2:%.*]] = call <1 x i64> @llvm.aarch64.neon.vsri.v1i64(<1 x i64> [[VSRI_N]], <1 x i64> [[VSRI_N1]], i32 33) 579 // CHECK: ret <1 x i64> [[VSRI_N2]] 580 poly64x1_t test_vsri_n_p64(poly64x1_t a, poly64x1_t b) { 581 return vsri_n_p64(a, b, 33); 582 } 583 584 // CHECK-LABEL: define <2 x i64> @test_vsriq_n_p64(<2 x i64> %a, <2 x i64> %b) #1 { 585 // CHECK: [[TMP0:%.*]] = bitcast <2 x i64> %a to <16 x i8> 586 // CHECK: [[TMP1:%.*]] = bitcast <2 x i64> %b to <16 x i8> 587 // CHECK: [[VSRI_N:%.*]] = bitcast <16 x i8> [[TMP0]] to <2 x i64> 588 // CHECK: [[VSRI_N1:%.*]] = bitcast <16 x i8> [[TMP1]] to <2 x i64> 589 // CHECK: [[VSRI_N2:%.*]] = call <2 x i64> @llvm.aarch64.neon.vsri.v2i64(<2 x i64> [[VSRI_N]], <2 x i64> [[VSRI_N1]], i32 64) 590 // CHECK: ret <2 x i64> [[VSRI_N2]] 591 poly64x2_t test_vsriq_n_p64(poly64x2_t a, poly64x2_t b) { 592 return vsriq_n_p64(a, b, 64); 593 } 594 595 // CHECK: attributes #0 ={{.*}}"min-legal-vector-width"="64" 596 // CHECK: attributes #1 ={{.*}}"min-legal-vector-width"="128" 597 // CHECK: attributes #2 ={{.*}}"min-legal-vector-width"="0" 598