Index: include/llvm/IR/IntrinsicsX86.td
===================================================================
--- include/llvm/IR/IntrinsicsX86.td
+++ include/llvm/IR/IntrinsicsX86.td
@@ -3860,6 +3860,18 @@
           [llvm_v16f32_ty, llvm_v16f32_ty, llvm_v16f32_ty,  llvm_i16_ty,
           llvm_i32_ty], [IntrNoMem]>;
 
+  def int_x86_avx512_vfmaddsd3_round :
+         GCCBuiltin<"__builtin_ia32_vfmaddsd3_round">,
+          Intrinsic<[llvm_v2f64_ty],
+          [llvm_v2f64_ty, llvm_v2f64_ty, llvm_v2f64_ty,  llvm_i8_ty,
+          llvm_i32_ty], [IntrNoMem]>;
+
+  def int_x86_avx512_vfmaddss3_round :
+         GCCBuiltin<"__builtin_ia32_vfmaddss3_round">,
+          Intrinsic<[llvm_v4f32_ty],
+          [llvm_v4f32_ty, llvm_v4f32_ty, llvm_v4f32_ty,  llvm_i8_ty,
+          llvm_i32_ty], [IntrNoMem]>;
+
   def int_x86_avx512_mask3_vfmsub_pd_128 :
          GCCBuiltin<"__builtin_ia32_vfmsubpd128_mask3">,
           Intrinsic<[llvm_v2f64_ty],
Index: lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- lib/Target/X86/X86ISelLowering.cpp
+++ lib/Target/X86/X86ISelLowering.cpp
@@ -16819,6 +16819,7 @@
                                               Src2, Src1, Src3),
                                   Mask, PassThru, Subtarget, DAG);
     }
+    case FMA_OP_MASK_SCA_RM:
     case FMA_OP_MASK3:
     case FMA_OP_MASKZ:
     case FMA_OP_MASK: {
@@ -16841,6 +16842,11 @@
       // First, we check if the intrinsic may have non-default rounding mode,
       // (IntrData->Opc1 != 0), then we check the rounding mode operand.
       unsigned IntrWithRoundingModeOpcode = IntrData->Opc1;
+      if (IntrData->Type == FMA_OP_MASK_SCA_RM)
+        return getScalarMaskingNode(DAG.getNode(IntrWithRoundingModeOpcode, dl,
+                                                Op.getValueType(), Src1, Src2,
+                                                Src3, Op.getOperand(5)),
+                                    Mask, PassThru, Subtarget, DAG);
       if (IntrWithRoundingModeOpcode != 0) {
         SDValue Rnd = Op.getOperand(5);
         if (cast<ConstantSDNode>(Rnd)->getZExtValue() !=
Index: lib/Target/X86/X86InstrAVX512.td
===================================================================
--- lib/Target/X86/X86InstrAVX512.td
+++ lib/Target/X86/X86InstrAVX512.td
@@ -4645,7 +4645,7 @@
                                                                   string SUFF> {
 
   defm NAME#213#SUFF: avx512_fma3s_common<opc213, OpcodeStr#"213"#_.Suffix , _ ,
-                (_.VT (OpNode _.RC:$src2, _.RC:$src1, _.RC:$src3)),
+                (_.VT (OpNodeRnd _.RC:$src2, _.RC:$src1, _.RC:$src3, (i32 FROUND_CURRENT))),
                 (_.VT (OpNode _.RC:$src2, _.RC:$src1,
                          (_.VT (scalar_to_vector(_.ScalarLdFrag addr:$src3))))),
                 (_.VT ( OpNodeRnd _.RC:$src2, _.RC:$src1, _.RC:$src3,
@@ -4656,7 +4656,7 @@
                          (_.ScalarLdFrag addr:$src3))))>;
 
   defm NAME#231#SUFF: avx512_fma3s_common<opc231, OpcodeStr#"231"#_.Suffix , _ ,
-                (_.VT (OpNode _.RC:$src2, _.RC:$src3, _.RC:$src1)),
+                (_.VT (OpNodeRnd _.RC:$src2, _.RC:$src3, _.RC:$src1, (i32 FROUND_CURRENT))),
                 (_.VT (OpNode _.RC:$src2,
                        (_.VT (scalar_to_vector(_.ScalarLdFrag addr:$src3))),
                               _.RC:$src1)),
@@ -4668,7 +4668,7 @@
                             (_.ScalarLdFrag addr:$src3), _.FRC:$src1)))>;
 
   defm NAME#132#SUFF: avx512_fma3s_common<opc132, OpcodeStr#"132"#_.Suffix , _ ,
-                (_.VT (OpNode _.RC:$src1, _.RC:$src3, _.RC:$src2)),
+                (_.VT (OpNodeRnd _.RC:$src1, _.RC:$src3, _.RC:$src2, (i32 FROUND_CURRENT))),
                 (_.VT (OpNode _.RC:$src1,
                        (_.VT (scalar_to_vector(_.ScalarLdFrag addr:$src3))),
                               _.RC:$src2)),
Index: lib/Target/X86/X86IntrinsicsInfo.h
===================================================================
--- lib/Target/X86/X86IntrinsicsInfo.h
+++ lib/Target/X86/X86IntrinsicsInfo.h
@@ -24,7 +24,7 @@
   INTR_TYPE_1OP_MASK, INTR_TYPE_1OP_MASK_RM,
   INTR_TYPE_2OP_MASK, INTR_TYPE_2OP_MASK_RM, INTR_TYPE_2OP_IMM8_MASK,
   INTR_TYPE_3OP_MASK, INTR_TYPE_3OP_MASK_RM, INTR_TYPE_3OP_IMM8_MASK,
-  FMA_OP_MASK, FMA_OP_MASKZ, FMA_OP_MASK3, VPERM_3OP_MASK,
+  FMA_OP_MASK, FMA_OP_MASK_SCA_RM,FMA_OP_MASKZ, FMA_OP_MASK3, VPERM_3OP_MASK,
   VPERM_3OP_MASKZ, INTR_TYPE_SCALAR_MASK,
   INTR_TYPE_SCALAR_MASK_RM, INTR_TYPE_3OP_SCALAR_MASK_RM,
   COMPRESS_EXPAND_IN_REG, COMPRESS_TO_MEM, BRCST_SUBVEC_TO_VEC,
@@ -2056,6 +2056,10 @@
   X86_INTRINSIC_DATA(avx512_rsqrt28_ss, INTR_TYPE_SCALAR_MASK_RM,X86ISD::RSQRT28, 0),
   X86_INTRINSIC_DATA(avx512_vcomi_sd, COMI_RM, X86ISD::COMI, X86ISD::UCOMI),
   X86_INTRINSIC_DATA(avx512_vcomi_ss, COMI_RM, X86ISD::COMI, X86ISD::UCOMI),
+  X86_INTRINSIC_DATA(avx512_vfmaddsd3_round, FMA_OP_MASK_SCA_RM, X86ISD::FMADD,
+                     X86ISD::FMADD_RND),
+  X86_INTRINSIC_DATA(avx512_vfmaddss3_round, FMA_OP_MASK_SCA_RM, X86ISD::FMADD,
+                     X86ISD::FMADD_RND),
   X86_INTRINSIC_DATA(avx_hadd_pd_256,   INTR_TYPE_2OP, X86ISD::FHADD, 0),
   X86_INTRINSIC_DATA(avx_hadd_ps_256,   INTR_TYPE_2OP, X86ISD::FHADD, 0),
   X86_INTRINSIC_DATA(avx_hsub_pd_256,   INTR_TYPE_2OP, X86ISD::FHSUB, 0),
Index: test/CodeGen/X86/avx512-intrinsics.ll
===================================================================
--- test/CodeGen/X86/avx512-intrinsics.ll
+++ test/CodeGen/X86/avx512-intrinsics.ll
@@ -7308,3 +7308,61 @@
   call void @llvm.x86.avx512.storent.ps.512(i8* %ptr, <16 x float> %data)
   ret void
 }
+
+declare <2 x double> @llvm.x86.avx512.vfmaddsd3.round(<2 x double>, <2 x double>, <2 x double>, i8, i32)
+
+define <2 x double>@test_int_x86_avx512_vfmaddsd3_round(<2 x double> %x0, <2 x double> %x1, <2 x double> %x2, i8 %x3,i32 %x4 ){
+; CHECK-LABEL: test_int_x86_avx512_vfmaddsd3_round:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    andl $1, %edi
+; CHECK-NEXT:    kmovw %edi, %k1
+; CHECK-NEXT:    vmovaps %zmm0, %zmm3
+; CHECK-NEXT:    vfmadd132sd %xmm1, %xmm2, %xmm3 {%k1}
+; CHECK-NEXT:    vmovaps %zmm1, %zmm4
+; CHECK-NEXT:    vfmadd213sd %xmm2, %xmm0, %xmm4
+; CHECK-NEXT:    vmovaps %zmm0, %zmm5
+; CHECK-NEXT:    vfmadd132sd {rz-sae}, %xmm1, %xmm2, %xmm5 {%k1}
+; CHECK-NEXT:    vfmadd213sd {rz-sae}, %xmm2, %xmm0, %xmm1
+; CHECK-NEXT:    vaddpd %xmm3, %xmm4, %xmm0
+; CHECK-NEXT:    vaddpd %xmm5, %xmm1, %xmm1
+; CHECK-NEXT:    vaddpd %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <2 x double> @llvm.x86.avx512.vfmaddsd3.round(<2 x double> %x0, <2 x double> %x1, <2 x double> %x2, i8 -1, i32 4)
+  %res1 = call <2 x double> @llvm.x86.avx512.vfmaddsd3.round(<2 x double> %x0, <2 x double> %x1, <2 x double> %x2, i8 %x3, i32 4)
+  %res2 = call <2 x double> @llvm.x86.avx512.vfmaddsd3.round(<2 x double> %x0, <2 x double> %x1, <2 x double> %x2, i8 -1, i32 3)
+  %res3 = call <2 x double> @llvm.x86.avx512.vfmaddsd3.round(<2 x double> %x0, <2 x double> %x1, <2 x double> %x2, i8 %x3, i32 3)
+  %res4 = fadd <2 x double> %res, %res1
+  %res5 = fadd <2 x double> %res2, %res3
+  %res6 = fadd <2 x double> %res4, %res5
+  ret <2 x double> %res6
+}
+
+declare <4 x float> @llvm.x86.avx512.vfmaddss3.round(<4 x float>, <4 x float>, <4 x float>, i8, i32)
+
+define <4 x float>@test_int_x86_avx512_vfmaddss3_round(<4 x float> %x0, <4 x float> %x1, <4 x float> %x2, i8 %x3,i32 %x4 ){
+; CHECK-LABEL: test_int_x86_avx512_vfmaddss3_round:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    andl $1, %edi
+; CHECK-NEXT:    kmovw %edi, %k1
+; CHECK-NEXT:    vmovaps %zmm0, %zmm3
+; CHECK-NEXT:    vfmadd132ss %xmm1, %xmm2, %xmm3 {%k1}
+; CHECK-NEXT:    vmovaps %zmm1, %zmm4
+; CHECK-NEXT:    vfmadd213ss %xmm2, %xmm0, %xmm4
+; CHECK-NEXT:    vmovaps %zmm0, %zmm5
+; CHECK-NEXT:    vfmadd132ss {rz-sae}, %xmm1, %xmm2, %xmm5 {%k1}
+; CHECK-NEXT:    vfmadd213ss {rz-sae}, %xmm2, %xmm0, %xmm1
+; CHECK-NEXT:    vaddps %xmm3, %xmm4, %xmm0
+; CHECK-NEXT:    vaddps %xmm5, %xmm1, %xmm1
+; CHECK-NEXT:    vaddps %xmm1, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <4 x float> @llvm.x86.avx512.vfmaddss3.round(<4 x float> %x0, <4 x float> %x1, <4 x float> %x2, i8 -1, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.vfmaddss3.round(<4 x float> %x0, <4 x float> %x1, <4 x float> %x2, i8 %x3, i32 4)
+  %res2 = call <4 x float> @llvm.x86.avx512.vfmaddss3.round(<4 x float> %x0, <4 x float> %x1, <4 x float> %x2, i8 -1, i32 3)
+  %res3 = call <4 x float> @llvm.x86.avx512.vfmaddss3.round(<4 x float> %x0, <4 x float> %x1, <4 x float> %x2, i8 %x3, i32 3)
+  %res4 = fadd <4 x float> %res, %res1
+  %res5 = fadd <4 x float> %res2, %res3
+  %res6 = fadd <4 x float> %res4, %res5
+  ret <4 x float> %res6
+}
+
+