[AVX-512] Remove 128-bit and 256-bit masked floating point add/sub/mul/div intrinsics and upgrade to native IR.

llvm-svn: 280633

5 files changed, 296 insertions, 316 deletions

diff --git a/llvm/include/llvm/IR/IntrinsicsX86.td b/llvm/include/llvm/IR/IntrinsicsX86.td
index 316749f9a6e..ad05850c56e 100644
--- a/llvm/include/llvm/IR/IntrinsicsX86.td
+++ b/llvm/include/llvm/IR/IntrinsicsX86.td
@@ -5233,75 +5233,27 @@ let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
 // Arithmetic ops
 let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
 
-  def int_x86_avx512_mask_add_ps_128 : GCCBuiltin<"__builtin_ia32_addps128_mask">,
-          Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_v4f32_ty,
-                     llvm_v4f32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_add_ps_256 : GCCBuiltin<"__builtin_ia32_addps256_mask">,
-          Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_v8f32_ty,
-                     llvm_v8f32_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_add_ps_512 : GCCBuiltin<"__builtin_ia32_addps512_mask">,
           Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_v16f32_ty,
                      llvm_v16f32_ty, llvm_i16_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_add_pd_128 : GCCBuiltin<"__builtin_ia32_addpd128_mask">,
-          Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty, llvm_v2f64_ty,
-                     llvm_v2f64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_add_pd_256 : GCCBuiltin<"__builtin_ia32_addpd256_mask">,
-          Intrinsic<[llvm_v4f64_ty], [llvm_v4f64_ty, llvm_v4f64_ty,
-                     llvm_v4f64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_add_pd_512 : GCCBuiltin<"__builtin_ia32_addpd512_mask">,
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_v8f64_ty,
                      llvm_v8f64_ty, llvm_i8_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_sub_ps_128 : GCCBuiltin<"__builtin_ia32_subps128_mask">,
-          Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_v4f32_ty,
-                     llvm_v4f32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_sub_ps_256 : GCCBuiltin<"__builtin_ia32_subps256_mask">,
-          Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_v8f32_ty,
-                     llvm_v8f32_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_sub_ps_512 : GCCBuiltin<"__builtin_ia32_subps512_mask">,
           Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_v16f32_ty,
                      llvm_v16f32_ty, llvm_i16_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_sub_pd_128 : GCCBuiltin<"__builtin_ia32_subpd128_mask">,
-          Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty, llvm_v2f64_ty,
-                     llvm_v2f64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_sub_pd_256 : GCCBuiltin<"__builtin_ia32_subpd256_mask">,
-          Intrinsic<[llvm_v4f64_ty], [llvm_v4f64_ty, llvm_v4f64_ty,
-                     llvm_v4f64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_sub_pd_512 : GCCBuiltin<"__builtin_ia32_subpd512_mask">,
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_v8f64_ty,
                      llvm_v8f64_ty, llvm_i8_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_mul_ps_128 : GCCBuiltin<"__builtin_ia32_mulps_mask">,
-          Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_v4f32_ty,
-                     llvm_v4f32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_mul_ps_256 : GCCBuiltin<"__builtin_ia32_mulps256_mask">,
-          Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_v8f32_ty,
-                     llvm_v8f32_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_mul_ps_512 : GCCBuiltin<"__builtin_ia32_mulps512_mask">,
           Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_v16f32_ty,
                      llvm_v16f32_ty, llvm_i16_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_mul_pd_128 : GCCBuiltin<"__builtin_ia32_mulpd_mask">,
-          Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty, llvm_v2f64_ty,
-                     llvm_v2f64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_mul_pd_256 : GCCBuiltin<"__builtin_ia32_mulpd256_mask">,
-          Intrinsic<[llvm_v4f64_ty], [llvm_v4f64_ty, llvm_v4f64_ty,
-                     llvm_v4f64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_mul_pd_512 : GCCBuiltin<"__builtin_ia32_mulpd512_mask">,
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_v8f64_ty,
                      llvm_v8f64_ty, llvm_i8_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_div_ps_128 : GCCBuiltin<"__builtin_ia32_divps_mask">,
-          Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty, llvm_v4f32_ty,
-                     llvm_v4f32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_div_ps_256 : GCCBuiltin<"__builtin_ia32_divps256_mask">,
-          Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_v8f32_ty,
-                     llvm_v8f32_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_div_ps_512 : GCCBuiltin<"__builtin_ia32_divps512_mask">,
           Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_v16f32_ty,
                      llvm_v16f32_ty, llvm_i16_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_div_pd_128 : GCCBuiltin<"__builtin_ia32_divpd_mask">,
-          Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty, llvm_v2f64_ty,
-                     llvm_v2f64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_div_pd_256 : GCCBuiltin<"__builtin_ia32_divpd256_mask">,
-          Intrinsic<[llvm_v4f64_ty], [llvm_v4f64_ty, llvm_v4f64_ty,
-                     llvm_v4f64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_div_pd_512 : GCCBuiltin<"__builtin_ia32_divpd512_mask">,
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_v8f64_ty,
                      llvm_v8f64_ty, llvm_i8_ty, llvm_i32_ty], [IntrNoMem]>;
diff --git a/llvm/lib/IR/AutoUpgrade.cpp b/llvm/lib/IR/AutoUpgrade.cpp
index 5f62802654e..79a8d18126d 100644
--- a/llvm/lib/IR/AutoUpgrade.cpp
+++ b/llvm/lib/IR/AutoUpgrade.cpp
@@ -274,6 +274,22 @@ static bool UpgradeIntrinsicFunction1(Function *F, Function *&NewFn) {
          Name.startswith("avx512.mask.padd.") ||
          Name.startswith("avx512.mask.psub.") ||
          Name.startswith("avx512.mask.pmull.") ||
+         Name.startswith("avx512.mask.add.pd.128") ||
+         Name.startswith("avx512.mask.add.pd.256") ||
+         Name.startswith("avx512.mask.add.ps.128") ||
+         Name.startswith("avx512.mask.add.ps.256") ||
+         Name.startswith("avx512.mask.div.pd.128") ||
+         Name.startswith("avx512.mask.div.pd.256") ||
+         Name.startswith("avx512.mask.div.ps.128") ||
+         Name.startswith("avx512.mask.div.ps.256") ||
+         Name.startswith("avx512.mask.mul.pd.128") ||
+         Name.startswith("avx512.mask.mul.pd.256") ||
+         Name.startswith("avx512.mask.mul.ps.128") ||
+         Name.startswith("avx512.mask.mul.ps.256") ||
+         Name.startswith("avx512.mask.sub.pd.128") ||
+         Name.startswith("avx512.mask.sub.pd.256") ||
+         Name.startswith("avx512.mask.sub.ps.128") ||
+         Name.startswith("avx512.mask.sub.ps.256") ||
          Name.startswith("sse41.pmovsx") ||
          Name.startswith("sse41.pmovzx") ||
          Name.startswith("avx2.pmovsx") ||
@@ -1249,6 +1265,34 @@ void llvm::UpgradeIntrinsicCall(CallInst *CI, Function *NewFn) {
       Rep = Builder.CreateMul(CI->getArgOperand(0), CI->getArgOperand(1));
       Rep = EmitX86Select(Builder, CI->getArgOperand(3), Rep,
                           CI->getArgOperand(2));
+    } else if (IsX86 && (Name.startswith("avx512.mask.add.pd.128") ||
+                         Name.startswith("avx512.mask.add.pd.256") ||
+                         Name.startswith("avx512.mask.add.ps.128") ||
+                         Name.startswith("avx512.mask.add.ps.256"))) {
+      Rep = Builder.CreateFAdd(CI->getArgOperand(0), CI->getArgOperand(1));
+      Rep = EmitX86Select(Builder, CI->getArgOperand(3), Rep,
+                          CI->getArgOperand(2));
+    } else if (IsX86 && (Name.startswith("avx512.mask.div.pd.128") ||
+                         Name.startswith("avx512.mask.div.pd.256") ||
+                         Name.startswith("avx512.mask.div.ps.128") ||
+                         Name.startswith("avx512.mask.div.ps.256"))) {
+      Rep = Builder.CreateFDiv(CI->getArgOperand(0), CI->getArgOperand(1));
+      Rep = EmitX86Select(Builder, CI->getArgOperand(3), Rep,
+                          CI->getArgOperand(2));
+    } else if (IsX86 && (Name.startswith("avx512.mask.mul.pd.128") ||
+                         Name.startswith("avx512.mask.mul.pd.256") ||
+                         Name.startswith("avx512.mask.mul.ps.128") ||
+                         Name.startswith("avx512.mask.mul.ps.256"))) {
+      Rep = Builder.CreateFMul(CI->getArgOperand(0), CI->getArgOperand(1));
+      Rep = EmitX86Select(Builder, CI->getArgOperand(3), Rep,
+                          CI->getArgOperand(2));
+    } else if (IsX86 && (Name.startswith("avx512.mask.sub.pd.128") ||
+                         Name.startswith("avx512.mask.sub.pd.256") ||
+                         Name.startswith("avx512.mask.sub.ps.128") ||
+                         Name.startswith("avx512.mask.sub.ps.256"))) {
+      Rep = Builder.CreateFSub(CI->getArgOperand(0), CI->getArgOperand(1));
+      Rep = EmitX86Select(Builder, CI->getArgOperand(3), Rep,
+                          CI->getArgOperand(2));
     } else {
       llvm_unreachable("Unknown function for CallInst upgrade.");
     }
diff --git a/llvm/lib/Target/X86/X86IntrinsicsInfo.h b/llvm/lib/Target/X86/X86IntrinsicsInfo.h
index c383f313b1c..11ca28ac321 100644
--- a/llvm/lib/Target/X86/X86IntrinsicsInfo.h
+++ b/llvm/lib/Target/X86/X86IntrinsicsInfo.h
@@ -378,12 +378,8 @@ static const IntrinsicData  IntrinsicsWithoutChain[] = {
   X86_INTRINSIC_DATA(avx512_kunpck_dq, KUNPCK, ISD::CONCAT_VECTORS, 0),
   X86_INTRINSIC_DATA(avx512_kunpck_wd, KUNPCK, ISD::CONCAT_VECTORS, 0),
 
-  X86_INTRINSIC_DATA(avx512_mask_add_pd_128, INTR_TYPE_2OP_MASK, ISD::FADD, 0),
-  X86_INTRINSIC_DATA(avx512_mask_add_pd_256, INTR_TYPE_2OP_MASK, ISD::FADD, 0),
   X86_INTRINSIC_DATA(avx512_mask_add_pd_512, INTR_TYPE_2OP_MASK, ISD::FADD,
   X86ISD::FADD_RND),
-  X86_INTRINSIC_DATA(avx512_mask_add_ps_128, INTR_TYPE_2OP_MASK, ISD::FADD, 0),
-  X86_INTRINSIC_DATA(avx512_mask_add_ps_256, INTR_TYPE_2OP_MASK, ISD::FADD, 0),
   X86_INTRINSIC_DATA(avx512_mask_add_ps_512, INTR_TYPE_2OP_MASK, ISD::FADD,
   X86ISD::FADD_RND),
   X86_INTRINSIC_DATA(avx512_mask_add_sd_round, INTR_TYPE_SCALAR_MASK_RM, ISD::FADD,
@@ -650,12 +646,8 @@ static const IntrinsicData  IntrinsicsWithoutChain[] = {
                      X86ISD::DBPSADBW, 0),
   X86_INTRINSIC_DATA(avx512_mask_dbpsadbw_512, INTR_TYPE_3OP_IMM8_MASK,
                      X86ISD::DBPSADBW, 0),
-  X86_INTRINSIC_DATA(avx512_mask_div_pd_128, INTR_TYPE_2OP_MASK, ISD::FDIV, 0),
-  X86_INTRINSIC_DATA(avx512_mask_div_pd_256, INTR_TYPE_2OP_MASK, ISD::FDIV, 0),
   X86_INTRINSIC_DATA(avx512_mask_div_pd_512, INTR_TYPE_2OP_MASK, ISD::FDIV,
                      X86ISD::FDIV_RND),
-  X86_INTRINSIC_DATA(avx512_mask_div_ps_128, INTR_TYPE_2OP_MASK, ISD::FDIV, 0),
-  X86_INTRINSIC_DATA(avx512_mask_div_ps_256, INTR_TYPE_2OP_MASK, ISD::FDIV, 0),
   X86_INTRINSIC_DATA(avx512_mask_div_ps_512, INTR_TYPE_2OP_MASK, ISD::FDIV,
                      X86ISD::FDIV_RND),
   X86_INTRINSIC_DATA(avx512_mask_div_sd_round, INTR_TYPE_SCALAR_MASK_RM, ISD::FDIV,
@@ -798,12 +790,8 @@ static const IntrinsicData  IntrinsicsWithoutChain[] = {
                      X86ISD::MOVSD, 0),
   X86_INTRINSIC_DATA(avx512_mask_move_ss, INTR_TYPE_SCALAR_MASK,
                      X86ISD::MOVSS, 0),
-  X86_INTRINSIC_DATA(avx512_mask_mul_pd_128, INTR_TYPE_2OP_MASK, ISD::FMUL, 0),
-  X86_INTRINSIC_DATA(avx512_mask_mul_pd_256, INTR_TYPE_2OP_MASK, ISD::FMUL, 0),
   X86_INTRINSIC_DATA(avx512_mask_mul_pd_512, INTR_TYPE_2OP_MASK, ISD::FMUL,
                      X86ISD::FMUL_RND),
-  X86_INTRINSIC_DATA(avx512_mask_mul_ps_128, INTR_TYPE_2OP_MASK, ISD::FMUL, 0),
-  X86_INTRINSIC_DATA(avx512_mask_mul_ps_256, INTR_TYPE_2OP_MASK, ISD::FMUL, 0),
   X86_INTRINSIC_DATA(avx512_mask_mul_ps_512, INTR_TYPE_2OP_MASK, ISD::FMUL,
                      X86ISD::FMUL_RND),
   X86_INTRINSIC_DATA(avx512_mask_mul_sd_round, INTR_TYPE_SCALAR_MASK_RM, ISD::FMUL,
@@ -1388,12 +1376,8 @@ static const IntrinsicData  IntrinsicsWithoutChain[] = {
                      X86ISD::FSQRT_RND, 0),
   X86_INTRINSIC_DATA(avx512_mask_sqrt_ss, INTR_TYPE_SCALAR_MASK_RM,
                      X86ISD::FSQRT_RND, 0),
-  X86_INTRINSIC_DATA(avx512_mask_sub_pd_128, INTR_TYPE_2OP_MASK, ISD::FSUB, 0),
-  X86_INTRINSIC_DATA(avx512_mask_sub_pd_256, INTR_TYPE_2OP_MASK, ISD::FSUB, 0),
   X86_INTRINSIC_DATA(avx512_mask_sub_pd_512, INTR_TYPE_2OP_MASK, ISD::FSUB,
                      X86ISD::FSUB_RND),
-  X86_INTRINSIC_DATA(avx512_mask_sub_ps_128, INTR_TYPE_2OP_MASK, ISD::FSUB, 0),
-  X86_INTRINSIC_DATA(avx512_mask_sub_ps_256, INTR_TYPE_2OP_MASK, ISD::FSUB, 0),
   X86_INTRINSIC_DATA(avx512_mask_sub_ps_512, INTR_TYPE_2OP_MASK, ISD::FSUB,
                      X86ISD::FSUB_RND),
   X86_INTRINSIC_DATA(avx512_mask_sub_sd_round, INTR_TYPE_SCALAR_MASK_RM, ISD::FSUB,
diff --git a/llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll b/llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
index 52a8099b7a1..3a7f46a8f2c 100644
--- a/llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
+++ b/llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
@@ -2935,3 +2935,251 @@ define <8 x i32> @test_mask_add_epi32_rmbkz_256(<8 x i32> %a, i32* %ptr_b, i8 %m
 
 declare <8 x i32> @llvm.x86.avx512.mask.padd.d.256(<8 x i32>, <8 x i32>, <8 x i32>, i8)
 
+define <8 x float> @test_mm512_maskz_add_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_add_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x58,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_mask_add_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_add_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x58,0xd1]
+; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_add_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_add_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x58,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
+  ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
+
+define <4 x float> @test_mm512_maskz_add_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_add_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x58,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_mask_add_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_add_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x58,0xd1]
+; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_add_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_add_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x58,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
+  ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
+
+define <8 x float> @test_mm512_maskz_sub_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_sub_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x5c,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_mask_sub_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_sub_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x5c,0xd1]
+; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_sub_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_sub_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x5c,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
+  ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
+
+define <4 x float> @test_mm512_maskz_sub_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_sub_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x5c,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_mask_sub_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_sub_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x5c,0xd1]
+; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_sub_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_sub_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x5c,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
+  ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
+
+define <8 x float> @test_mm512_maskz_mul_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_mul_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x59,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_mask_mul_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_mul_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x59,0xd1]
+; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_mul_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_mul_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x59,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
+  ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
+
+define <4 x float> @test_mm512_maskz_mul_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_mul_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x59,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_mask_mul_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_mul_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x59,0xd1]
+; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_mul_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_mul_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x59,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
+  ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
+
+define <8 x float> @test_mm512_maskz_div_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_div_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x5e,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_mask_div_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_div_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x5e,0xd1]
+; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
+  ret <8 x float> %res
+}
+
+define <8 x float> @test_mm512_div_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_div_ps_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x5e,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
+  ret <8 x float> %res
+}
+declare <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
+
+define <4 x float> @test_mm512_maskz_div_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_maskz_div_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x5e,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_mask_div_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
+; CHECK-LABEL: test_mm512_mask_div_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
+; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x5e,0xd1]
+; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
+  ret <4 x float> %res
+}
+
+define <4 x float> @test_mm512_div_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
+; CHECK-LABEL: test_mm512_div_ps_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x5e,0xc1]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
+  ret <4 x float> %res
+}
+declare <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
+
diff --git a/llvm/test/CodeGen/X86/avx512vl-intrinsics.ll b/llvm/test/CodeGen/X86/avx512vl-intrinsics.ll
index 6d35c828dcc..9a913186315 100644
--- a/llvm/test/CodeGen/X86/avx512vl-intrinsics.ll
+++ b/llvm/test/CodeGen/X86/avx512vl-intrinsics.ll
@@ -1483,254 +1483,6 @@ define i8 @test_cmppd_128(<2 x double> %a, <2 x double> %b) {
  }
  declare i8 @llvm.x86.avx512.mask.cmp.pd.128(<2 x double> , <2 x double> , i32, i8)
 
-define <8 x float> @test_mm512_maskz_add_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_add_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x58,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_mask_add_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_add_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x58,0xd1]
-; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_add_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_add_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vaddps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x58,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
-  ret <8 x float> %res
-}
-declare <8 x float> @llvm.x86.avx512.mask.add.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
-
-define <4 x float> @test_mm512_maskz_add_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_add_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x58,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_mask_add_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_add_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x58,0xd1]
-; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_add_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_add_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x58,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
-  ret <4 x float> %res
-}
-declare <4 x float> @llvm.x86.avx512.mask.add.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
-
-define <8 x float> @test_mm512_maskz_sub_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_sub_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x5c,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_mask_sub_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_sub_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x5c,0xd1]
-; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_sub_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_sub_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vsubps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x5c,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
-  ret <8 x float> %res
-}
-declare <8 x float> @llvm.x86.avx512.mask.sub.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
-
-define <4 x float> @test_mm512_maskz_sub_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_sub_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x5c,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_mask_sub_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_sub_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x5c,0xd1]
-; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_sub_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_sub_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vsubps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x5c,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
-  ret <4 x float> %res
-}
-declare <4 x float> @llvm.x86.avx512.mask.sub.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
-
-define <8 x float> @test_mm512_maskz_mul_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_mul_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x59,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_mask_mul_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_mul_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x59,0xd1]
-; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_mul_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_mul_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vmulps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x59,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
-  ret <8 x float> %res
-}
-declare <8 x float> @llvm.x86.avx512.mask.mul.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
-
-define <4 x float> @test_mm512_maskz_mul_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_mul_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x59,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_mask_mul_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_mul_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x59,0xd1]
-; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_mul_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_mul_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vmulps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x59,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
-  ret <4 x float> %res
-}
-declare <4 x float> @llvm.x86.avx512.mask.mul.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
-
-define <8 x float> @test_mm512_maskz_div_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_div_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0xa9,0x5e,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_mask_div_ps_256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_div_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x29,0x5e,0xd1]
-; CHECK-NEXT:    vmovaps %ymm2, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %src, i8 %mask)
-  ret <8 x float> %res
-}
-
-define <8 x float> @test_mm512_div_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_div_ps_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vdivps %ymm1, %ymm0, %ymm0 ## encoding: [0x62,0xf1,0x7c,0x28,0x5e,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float>zeroinitializer, i8 -1)
-  ret <8 x float> %res
-}
-declare <8 x float> @llvm.x86.avx512.mask.div.ps.256(<8 x float>, <8 x float>, <8 x float>, i8)
-
-define <4 x float> @test_mm512_maskz_div_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_maskz_div_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm0 {%k1} {z} ## encoding: [0x62,0xf1,0x7c,0x89,0x5e,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_mask_div_ps_128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask) {
-; CHECK-LABEL: test_mm512_mask_div_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %edi, %k1 ## encoding: [0xc5,0xf8,0x92,0xcf]
-; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm2 {%k1} ## encoding: [0x62,0xf1,0x7c,0x09,0x5e,0xd1]
-; CHECK-NEXT:    vmovaps %xmm2, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x28,0xc2]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float> %src, i8 %mask)
-  ret <4 x float> %res
-}
-
-define <4 x float> @test_mm512_div_ps_128(<4 x float> %a0, <4 x float> %a1, i8 %mask) {
-; CHECK-LABEL: test_mm512_div_ps_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    vdivps %xmm1, %xmm0, %xmm0 ## encoding: [0x62,0xf1,0x7c,0x08,0x5e,0xc1]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float> %a0, <4 x float> %a1, <4 x float>zeroinitializer, i8 -1)
-  ret <4 x float> %res
-}
-declare <4 x float> @llvm.x86.avx512.mask.div.ps.128(<4 x float>, <4 x float>, <4 x float>, i8)
-
 define <8 x float> @test_mm512_maskz_max_ps_256(<8 x float> %a0, <8 x float> %a1, i8 %mask) {
 ; CHECK-LABEL: test_mm512_maskz_max_ps_256:
 ; CHECK:       ## BB#0:
@@ -5532,9 +5284,9 @@ define <8 x i32>@test_int_x86_avx512_mask_psrav8_si_const() {
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    vmovdqa32 {{.*#+}} ymm0 = [2,9,4294967284,23,4294967270,37,4294967256,51]
 ; CHECK-NEXT:    ## encoding: [0x62,0xf1,0x7d,0x28,0x6f,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI335_0-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI311_0-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    vpsravd {{.*}}(%rip), %ymm0, %ymm0 ## encoding: [0x62,0xf2,0x7d,0x28,0x46,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI335_1-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI311_1-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    retq ## encoding: [0xc3]
   %res = call <8 x i32> @llvm.x86.avx512.mask.psrav8.si(<8 x i32> <i32 2, i32 9, i32 -12, i32 23, i32 -26, i32 37, i32 -40, i32 51>, <8 x i32> <i32 1, i32 18, i32 35, i32 52, i32 69, i32 15, i32 32, i32 49>, <8 x i32> zeroinitializer, i8 -1)
   ret <8 x i32> %res
@@ -5565,9 +5317,9 @@ define <2 x i64>@test_int_x86_avx512_mask_psrav_q_128_const(i8 %x3) {
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    vmovdqa64 {{.*#+}} xmm0 = [2,18446744073709551607]
 ; CHECK-NEXT:    ## encoding: [0x62,0xf1,0xfd,0x08,0x6f,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI337_0-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI313_0-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    vpsravq {{.*}}(%rip), %xmm0, %xmm0 ## encoding: [0x62,0xf2,0xfd,0x08,0x46,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI337_1-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI313_1-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    retq ## encoding: [0xc3]
   %res = call <2 x i64> @llvm.x86.avx512.mask.psrav.q.128(<2 x i64> <i64 2, i64 -9>, <2 x i64> <i64 1, i64 90>, <2 x i64> zeroinitializer, i8 -1)
   ret <2 x i64> %res