fma-fneg-combine.ll 22.9 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -mattr=+avx512bw -mattr=+avx512vl -mattr=+avx512dq  | FileCheck %s  --check-prefix=CHECK --check-prefix=SKX
; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -mattr=+avx512f -mattr=+fma | FileCheck %s --check-prefix=CHECK --check-prefix=KNL

; This test checks combinations of FNEG and FMA intrinsics on AVX-512 target
; PR28892

define <16 x float> @test1(<16 x float> %a, <16 x float> %b, <16 x float> %c)  {
; CHECK-LABEL: test1:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmsub213ps {{.*#+}} zmm0 = (zmm1 * zmm0) - zmm2
; CHECK-NEXT:    retq
entry:
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %sub.i, i32 4) #2
  ret <16 x float> %0
}

declare <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float>, <16 x float>, <16 x float>, i32)
declare <16 x float> @llvm.x86.avx512.mask.vfnmadd.ps.512(<16 x float>, <16 x float>, <16 x float>, i16, i32)
declare <16 x float> @llvm.x86.avx512.mask.vfnmsub.ps.512(<16 x float>, <16 x float>, <16 x float>, i16, i32)


define <16 x float> @test2(<16 x float> %a, <16 x float> %b, <16 x float> %c) {
; CHECK-LABEL: test2:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfnmsub213ps {{.*#+}} zmm0 = -(zmm1 * zmm0) - zmm2
; CHECK-NEXT:    retq
entry:
  %0 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %c, i32 4) #2
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <16 x float> %sub.i
}

define <16 x float> @test3(<16 x float> %a, <16 x float> %b, <16 x float> %c)  {
; CHECK-LABEL: test3:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmsub213ps {{.*#+}} zmm0 = (zmm1 * zmm0) - zmm2
; CHECK-NEXT:    retq
entry:
  %0 = tail call <16 x float> @llvm.x86.avx512.mask.vfnmadd.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 -1, i32 4) #2
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <16 x float> %sub.i
}

define <16 x float> @test4(<16 x float> %a, <16 x float> %b, <16 x float> %c) {
; CHECK-LABEL: test4:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmadd213ps {{.*#+}} zmm0 = (zmm1 * zmm0) + zmm2
; CHECK-NEXT:    retq
entry:
  %0 = tail call <16 x float> @llvm.x86.avx512.mask.vfnmsub.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 -1, i32 4) #2
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <16 x float> %sub.i
}

define <16 x float> @test5(<16 x float> %a, <16 x float> %b, <16 x float> %c) {
; CHECK-LABEL: test5:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmsub213ps {ru-sae}, %zmm2, %zmm1, %zmm0
; CHECK-NEXT:    retq
entry:
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %sub.i, i32 10) #2
  ret <16 x float> %0
}

define <16 x float> @test6(<16 x float> %a, <16 x float> %b, <16 x float> %c) {
; CHECK-LABEL: test6:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmadd213ps {ru-sae}, %zmm2, %zmm1, %zmm0
; CHECK-NEXT:    retq
entry:
  %0 = tail call <16 x float> @llvm.x86.avx512.mask.vfnmsub.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 -1, i32 10) #2
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <16 x float> %sub.i
}


define <8 x float> @test7(<8 x float> %a, <8 x float> %b, <8 x float> %c) {
; CHECK-LABEL: test7:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfnmadd213ps {{.*#+}} ymm0 = -(ymm1 * ymm0) + ymm2
; CHECK-NEXT:    retq
entry:
  %0 = tail call <8 x float> @llvm.x86.fma.vfmsub.ps.256(<8 x float> %a, <8 x float> %b, <8 x float> %c) #2
  %sub.i = fsub <8 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <8 x float> %sub.i
}

define <8 x float> @test8(<8 x float> %a, <8 x float> %b, <8 x float> %c) {
; CHECK-LABEL: test8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmadd213ps {{.*#+}} ymm0 = (ymm1 * ymm0) + ymm2
; CHECK-NEXT:    retq
entry:
  %sub.c = fsub <8 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <8 x float> @llvm.x86.fma.vfmsub.ps.256(<8 x float> %a, <8 x float> %b, <8 x float> %sub.c) #2
  ret <8 x float> %0
}

declare <8 x float> @llvm.x86.fma.vfmsub.ps.256(<8 x float>, <8 x float>, <8 x float>)


define <8 x double> @test9(<8 x double> %a, <8 x double> %b, <8 x double> %c) {
; CHECK-LABEL: test9:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfnmsub213pd {{.*#+}} zmm0 = -(zmm1 * zmm0) - zmm2
; CHECK-NEXT:    retq
entry:
  %0 = tail call <8 x double> @llvm.x86.avx512.vfmadd.pd.512(<8 x double> %a, <8 x double> %b, <8 x double> %c, i32 4) #2
  %sub.i = fsub <8 x double> <double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00>, %0
  ret <8 x double> %sub.i
}

declare <8 x double> @llvm.x86.avx512.vfmadd.pd.512(<8 x double> %a, <8 x double> %b, <8 x double> %c, i32)

define <2 x double> @test10(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test10:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfmadd213sd {{.*#+}} xmm0 = (xmm1 * xmm0) + xmm2
; CHECK-NEXT:    vxorpd {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT:    retq
entry:
  %0 = tail call <2 x double> @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %a, <2 x double> %b, <2 x double> %c, i8 -1, i32 4) #2
  %sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %0
  ret <2 x double> %sub.i
}

declare <2 x double> @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %a, <2 x double> %b, <2 x double> %c, i8, i32)

define <4 x float> @test11(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test11:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    vxorps {{.*}}(%rip){1to4}, %xmm2, %xmm3
; SKX-NEXT:    vfmsub213ss {{.*#+}} xmm0 = (xmm1 * xmm0) - xmm2
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vmovss %xmm0, %xmm3, %xmm3 {%k1}
; SKX-NEXT:    vmovaps %xmm3, %xmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test11:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    vbroadcastss {{.*#+}} xmm3 = [-0.0E+0,-0.0E+0,-0.0E+0,-0.0E+0]
; KNL-NEXT:    vxorps %xmm3, %xmm2, %xmm3
; KNL-NEXT:    vfmsub213ss {{.*#+}} xmm0 = (xmm1 * xmm0) - xmm2
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vmovss %xmm0, %xmm3, %xmm3 {%k1}
; KNL-NEXT:    vmovaps %xmm3, %xmm0
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <4 x float> @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %sub.i, i8 %mask, i32 4) #10
  ret <4 x float> %0
}

declare <4 x float> @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float>, <4 x float>, <4 x float>, i8, i32)

define <4 x float> @test11b(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test11b:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfmsub213ss {{.*#+}} xmm0 {%k1} = (xmm1 * xmm0) - xmm2
; SKX-NEXT:    retq
;
; KNL-LABEL: test11b:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfmsub213ss {{.*#+}} xmm0 {%k1} = (xmm1 * xmm0) - xmm2
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %sub.i, i8 %mask, i32 4) #10
  ret <4 x float> %0
}

declare <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float>, <4 x float>, <4 x float>, i8, i32)

define <8 x double> @test12(<8 x double> %a, <8 x double> %b, <8 x double> %c, i8 %mask) {
; SKX-LABEL: test12:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfmadd132pd {{.*#+}} zmm0 {%k1} = (zmm0 * zmm1) + zmm2
; SKX-NEXT:    vxorpd {{.*}}(%rip){1to8}, %zmm0, %zmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test12:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfmadd132pd {{.*#+}} zmm0 {%k1} = (zmm0 * zmm1) + zmm2
; KNL-NEXT:    vpxorq {{.*}}(%rip){1to8}, %zmm0, %zmm0
; KNL-NEXT:    retq
entry:
  %0 = tail call <8 x double> @llvm.x86.avx512.vfmadd.pd.512(<8 x double> %a, <8 x double> %b, <8 x double> %c, i32 4) #2
  %bc = bitcast i8 %mask to <8 x i1>
  %sel = select <8 x i1> %bc, <8 x double> %0, <8 x double> %a
  %sub.i = fsub <8 x double> <double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00>, %sel
  ret <8 x double> %sub.i
}

define <2 x double> @test13(<2 x double> %a, <2 x double> %b, <2 x double> %c, i8 %mask) {
; SKX-LABEL: test13:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    vxorpd {{.*}}(%rip), %xmm0, %xmm3
; SKX-NEXT:    vfnmadd213sd {{.*#+}} xmm1 = -(xmm0 * xmm1) + xmm2
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vmovsd %xmm1, %xmm3, %xmm3 {%k1}
; SKX-NEXT:    vmovapd %xmm3, %xmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test13:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    vxorpd {{.*}}(%rip), %xmm0, %xmm3
; KNL-NEXT:    vfnmadd213sd {{.*#+}} xmm1 = -(xmm0 * xmm1) + xmm2
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vmovsd %xmm1, %xmm3, %xmm3 {%k1}
; KNL-NEXT:    vmovapd %xmm3, %xmm0
; KNL-NEXT:    retq

entry:
  %sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %a
  %0 = tail call <2 x double> @llvm.x86.avx512.mask.vfmadd.sd(<2 x double> %sub.i, <2 x double> %b, <2 x double> %c, i8 %mask, i32 4)
  ret <2 x double> %0
}

define <16 x float> @test14(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 %mask) {
; SKX-LABEL: test14:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmsub132ps {ru-sae}, %zmm1, %zmm2, %zmm0 {%k1}
; SKX-NEXT:    vxorps {{.*}}(%rip){1to16}, %zmm0, %zmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test14:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmsub132ps {ru-sae}, %zmm1, %zmm2, %zmm0 {%k1}
; KNL-NEXT:    vpxord {{.*}}(%rip){1to16}, %zmm0, %zmm0
; KNL-NEXT:    retq
entry:
  %0 = tail call <16 x float> @llvm.x86.avx512.mask.vfnmsub.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 %mask, i32 10) #2
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %0
  ret <16 x float> %sub.i
}

define <16 x float> @test15(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 %mask)  {
; SKX-LABEL: test15:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vxorps {{.*}}(%rip){1to16}, %zmm0, %zmm3
; SKX-NEXT:    vfnmadd213ps {ru-sae}, %zmm2, %zmm0, %zmm1
; SKX-NEXT:    vmovaps %zmm1, %zmm3 {%k1}
; SKX-NEXT:    vfnmadd132ps {rd-sae}, %zmm0, %zmm2, %zmm3 {%k1}
; SKX-NEXT:    vmovaps %zmm3, %zmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test15:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vpxord {{.*}}(%rip){1to16}, %zmm0, %zmm3
; KNL-NEXT:    vfnmadd213ps {ru-sae}, %zmm2, %zmm0, %zmm1
; KNL-NEXT:    vmovaps %zmm1, %zmm3 {%k1}
; KNL-NEXT:    vfnmadd132ps {rd-sae}, %zmm0, %zmm2, %zmm3 {%k1}
; KNL-NEXT:    vmovaps %zmm3, %zmm0
; KNL-NEXT:    retq
entry:
  %bc = bitcast i16 %mask to <16 x i1>
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %a
  %0 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %sub.i, <16 x float> %b, <16 x float> %c, i32 10)
  %sel = select <16 x i1> %bc, <16 x float> %0, <16 x float> %sub.i
  %1 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %sel, <16 x float> %sub.i, <16 x float> %c, i32 9)
  %sel2 = select <16 x i1> %bc, <16 x float> %1, <16 x float> %sel
  ret <16 x float> %sel2
}

define <16 x float> @test16(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 %mask) {
; SKX-LABEL: test16:
; SKX:       # %bb.0:
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfmsubadd132ps {rd-sae}, %zmm1, %zmm2, %zmm0 {%k1}
; SKX-NEXT:    retq
;
; KNL-LABEL: test16:
; KNL:       # %bb.0:
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfmsubadd132ps {rd-sae}, %zmm1, %zmm2, %zmm0 {%k1}
; KNL-NEXT:    retq
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %res = call <16 x float> @llvm.x86.avx512.vfmaddsub.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %sub.i, i32 9)
  %bc = bitcast i16 %mask to <16 x i1>
  %sel = select <16 x i1> %bc, <16 x float> %res, <16 x float> %a
  ret <16 x float> %sel
}
declare <16 x float> @llvm.x86.avx512.vfmaddsub.ps.512(<16 x float>, <16 x float>, <16 x float>, i32)

define <8 x double> @test17(<8 x double> %a, <8 x double> %b, <8 x double> %c, i8 %mask) {
; SKX-LABEL: test17:
; SKX:       # %bb.0:
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfmsubadd132pd {{.*#+}} zmm0 {%k1} = (zmm0 * zmm1) -/+ zmm2
; SKX-NEXT:    retq
;
; KNL-LABEL: test17:
; KNL:       # %bb.0:
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfmsubadd132pd {{.*#+}} zmm0 {%k1} = (zmm0 * zmm1) -/+ zmm2
; KNL-NEXT:    retq
  %sub.i = fsub <8 x double> <double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00>, %c
  %res = call <8 x double> @llvm.x86.avx512.vfmaddsub.pd.512(<8 x double> %a, <8 x double> %b, <8 x double> %sub.i, i32 4)
  %bc = bitcast i8 %mask to <8 x i1>
  %sel = select <8 x i1> %bc, <8 x double> %res, <8 x double> %a
  ret <8 x double> %sel
}
declare <8 x double> @llvm.x86.avx512.vfmaddsub.pd.512(<8 x double>, <8 x double>, <8 x double>, i32)

define <4 x float> @test18(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test18:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmadd213ss {{.*#+}} xmm0 {%k1} = -(xmm1 * xmm0) + xmm2
; SKX-NEXT:    retq
;
; KNL-LABEL: test18:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmadd213ss {{.*#+}} xmm0 {%k1} = -(xmm1 * xmm0) + xmm2
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %c, i8 %mask, i32 4) #10
  ret <4 x float> %0
}

define <4 x float> @test19(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test19:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmsub213ss {{.*#+}} xmm0 {%k1} = -(xmm1 * xmm0) - xmm2
; SKX-NEXT:    retq
;
; KNL-LABEL: test19:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmsub213ss {{.*#+}} xmm0 {%k1} = -(xmm1 * xmm0) - xmm2
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %sub.i.2 = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %sub.i.2, i8 %mask, i32 4) #10
  ret <4 x float> %0
}

define <4 x float> @test20(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test20:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmadd231ss {{.*#+}} xmm2 {%k1} = -(xmm0 * xmm1) + xmm2
; SKX-NEXT:    vmovaps %xmm2, %xmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test20:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmadd231ss {{.*#+}} xmm2 {%k1} = -(xmm0 * xmm1) + xmm2
; KNL-NEXT:    vmovaps %xmm2, %xmm0
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %0 = tail call <4 x float> @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %c, i8 %mask, i32 4) #10
  ret <4 x float> %0
}

define <4 x float> @test21(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test21:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmadd213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; SKX-NEXT:    retq
;
; KNL-LABEL: test21:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmadd213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %c, i8 %mask, i32 8) #10
  ret <4 x float> %0
}

define <4 x float> @test22(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test22:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmsub213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; SKX-NEXT:    retq
;
; KNL-LABEL: test22:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmsub213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %sub.i.2 = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %sub.i.2, i8 %mask, i32 8) #10
  ret <4 x float> %0
}

define <4 x float> @test23(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test23:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfnmadd231ss {rn-sae}, %xmm1, %xmm0, %xmm2 {%k1}
; SKX-NEXT:    vmovaps %xmm2, %xmm0
; SKX-NEXT:    retq
;
; KNL-LABEL: test23:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfnmadd231ss {rn-sae}, %xmm1, %xmm0, %xmm2 {%k1}
; KNL-NEXT:    vmovaps %xmm2, %xmm0
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %0 = tail call <4 x float> @llvm.x86.avx512.mask3.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %c, i8 %mask, i32 8) #10
  ret <4 x float> %0
}

define <4 x float> @test24(<4 x float> %a, <4 x float> %b, <4 x float> %c, i8 zeroext %mask) local_unnamed_addr #0 {
; SKX-LABEL: test24:
; SKX:       # %bb.0: # %entry
; SKX-NEXT:    kmovd %edi, %k1
; SKX-NEXT:    vfmsub213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; SKX-NEXT:    retq
;
; KNL-LABEL: test24:
; KNL:       # %bb.0: # %entry
; KNL-NEXT:    kmovw %edi, %k1
; KNL-NEXT:    vfmsub213ss {rn-sae}, %xmm2, %xmm1, %xmm0 {%k1}
; KNL-NEXT:    retq
entry:
  %sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <4 x float> @llvm.x86.avx512.mask.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %sub.i, i8 %mask, i32 8) #10
  ret <4 x float> %0
}

define <16 x float> @test25(<16 x float> %a, <16 x float> %b, <16 x float> %c)  {
; CHECK-LABEL: test25:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    vfnmsub213ps {rn-sae}, %zmm2, %zmm1, %zmm0
; CHECK-NEXT:    retq
entry:
  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
  %sub.i.2 = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
  %0 = tail call <16 x float> @llvm.x86.avx512.vfmadd.ps.512(<16 x float> %a, <16 x float> %sub.i, <16 x float> %sub.i.2, i32 8) #2
  ret <16 x float> %0
}