[RISCV][VP] Lower VP_FMA to RVV instructions.

[craig.topper]

Summary

We already had FMA_VL node, but we didn't have masked patterns.
I have not added the fneg variations. I'll do those after I add
llvm.vp.fneg.

[frasercrmck]

Now that D118810 landed, I think we can rebase this and switch true_mask patterns to use V0 and rely on the post-process step to select the unmasked FMAs. You'll need to add the mask information to the ternary pseudos, by subclassing them from
RISCVMaskedPseudo. Hopefully that should cut down on the net pattern gain.

[craig.topper]

In D119196#3307359, @frasercrmck wrote:

Now that D118810 landed, I think we can rebase this and switch true_mask patterns to use V0 and rely on the post-process step to select the unmasked FMAs. You'll need to add the mask information to the ternary pseudos, by subclassing them from
RISCVMaskedPseudo. Hopefully that should cut down on the net pattern gain.

The existing operand copying rules in the post-process don't work for FMAs. We need to retain the merge operand during the conversion and I'm not sure how to detect that yet.. Can I do it as a follow up?

[frasercrmck]

In D119196#3308046, @craig.topper wrote:

In D119196#3307359, @frasercrmck wrote:

Now that D118810 landed, I think we can rebase this and switch true_mask patterns to use V0 and rely on the post-process step to select the unmasked FMAs. You'll need to add the mask information to the ternary pseudos, by subclassing them from
RISCVMaskedPseudo. Hopefully that should cut down on the net pattern gain.

The existing operand copying rules in the post-process don't work for FMAs. We need to retain the merge operand during the conversion and I'm not sure how to detect that yet.. Can I do it as a follow up?

Oh yes I missed that: no worries. I didn't mean it to be a must-have. I'm sure we can tackle it later.

LGTM.

llvm/lib/Target/RISCV/RISCVInstrInfoVVLPatterns.td

1133

Hopefully this could be improved by D117670?

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

static const unsigned IntegerVPOps[] = {
         ISD::VP_XOR,         ISD::VP_ASHR,        ISD::VP_LSHR,
         ISD::VP_SHL,         ISD::VP_REDUCE_ADD,  ISD::VP_REDUCE_AND,
         ISD::VP_REDUCE_OR,   ISD::VP_REDUCE_XOR,  ISD::VP_REDUCE_SMAX,
         ISD::VP_REDUCE_SMIN, ISD::VP_REDUCE_UMAX, ISD::VP_REDUCE_UMIN,
         ISD::VP_MERGE,       ISD::VP_SELECT};
 
     static const unsigned FloatingPointVPOps[] = {
         ISD::VP_FADD,        ISD::VP_FSUB,        ISD::VP_FMUL,
-        ISD::VP_FDIV,        ISD::VP_FNEG,        ISD::VP_REDUCE_FADD,
-        ISD::VP_REDUCE_SEQ_FADD, ISD::VP_REDUCE_FMIN, ISD::VP_REDUCE_FMAX,
-        ISD::VP_MERGE,       ISD::VP_SELECT};
+        ISD::VP_FDIV,        ISD::VP_FNEG,        ISD::VP_FMA,
+        ISD::VP_REDUCE_FADD, ISD::VP_REDUCE_SEQ_FADD, ISD::VP_REDUCE_FMIN,
+        ISD::VP_REDUCE_FMAX, ISD::VP_MERGE,       ISD::VP_SELECT};
 
     if (!Subtarget.is64Bit()) {
       // We must custom-lower certain vXi64 operations on RV32 due to the vector
       // element type being illegal.
       setOperationAction(ISD::INSERT_VECTOR_ELT, MVT::i64, Custom);
       setOperationAction(ISD::EXTRACT_VECTOR_ELT, MVT::i64, Custom);
 
       setOperationAction(ISD::VECREDUCE_ADD, MVT::i64, Custom);
SDValue RISCVTargetLowering::LowerOperation(SDValue Op,
   case ISD::VP_FSUB:
     return lowerVPOp(Op, DAG, RISCVISD::FSUB_VL);
   case ISD::VP_FMUL:
     return lowerVPOp(Op, DAG, RISCVISD::FMUL_VL);
   case ISD::VP_FDIV:
     return lowerVPOp(Op, DAG, RISCVISD::FDIV_VL);
   case ISD::VP_FNEG:
     return lowerVPOp(Op, DAG, RISCVISD::FNEG_VL);
+  case ISD::VP_FMA:
+    return lowerVPOp(Op, DAG, RISCVISD::FMA_VL);
   }
 }
 
 static SDValue getTargetNode(GlobalAddressSDNode *N, SDLoc DL, EVT Ty,
                              SelectionDAG &DAG, unsigned Flags) {
   return DAG.getTargetGlobalAddress(N->getGlobal(), DL, Ty, 0, Flags);
 }
 

llvm/lib/Target/RISCV/RISCVInstrInfoVVLPatterns.td

foreach vti = AllFloatVectors in {
   defvar suffix = vti.LMul.MX;
   def : Pat<(vti.Vector (riscv_fma_vl vti.RegClass:$rs1, vti.RegClass:$rd,
                                       vti.RegClass:$rs2, (vti.Mask true_mask),
                                       VLOpFrag)),
             (!cast<Instruction>("PseudoVFMADD_VV_"# suffix)
                  vti.RegClass:$rd, vti.RegClass:$rs1, vti.RegClass:$rs2,
                  GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
   def : Pat<(vti.Vector (riscv_fma_vl vti.RegClass:$rs1, vti.RegClass:$rd,
+                                      vti.RegClass:$rs2, (vti.Mask V0),
+                                      VLOpFrag)),
+            (!cast<Instruction>("PseudoVFMADD_VV_"# suffix #"_MASK")
+                 vti.RegClass:$rd, vti.RegClass:$rs1, vti.RegClass:$rs2,
+                 (vti.Mask V0), GPR:$vl, vti.Log2SEW)>;
+
+  def : Pat<(vti.Vector (riscv_fma_vl vti.RegClass:$rs1, vti.RegClass:$rd,
                                       (riscv_fneg_vl vti.RegClass:$rs2,
                                                      (vti.Mask true_mask),
                                                      VLOpFrag),
                                       (vti.Mask true_mask),
                                       VLOpFrag)),
             (!cast<Instruction>("PseudoVFMSUB_VV_"# suffix)
                  vti.RegClass:$rd, vti.RegClass:$rs1, vti.RegClass:$rs2,
                  GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
+
   def : Pat<(vti.Vector (riscv_fma_vl (riscv_fneg_vl vti.RegClass:$rs1,
                                                      (vti.Mask true_mask),
                                                      VLOpFrag),
                                       vti.RegClass:$rd,
                                       (riscv_fneg_vl vti.RegClass:$rs2,
                                                      (vti.Mask true_mask),
                                                      VLOpFrag),
                                       (vti.Mask true_mask),
                                       VLOpFrag)),
             (!cast<Instruction>("PseudoVFNMADD_VV_"# suffix)
                  vti.RegClass:$rd, vti.RegClass:$rs1, vti.RegClass:$rs2,
                  GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
+
   def : Pat<(vti.Vector (riscv_fma_vl (riscv_fneg_vl vti.RegClass:$rs1,
                                                      (vti.Mask true_mask),
                                                      VLOpFrag),
                                       vti.RegClass:$rd, vti.RegClass:$rs2,
                                       (vti.Mask true_mask),
                                       VLOpFrag)),
             (!cast<Instruction>("PseudoVFNMSUB_VV_"# suffix)
                  vti.RegClass:$rd, vti.RegClass:$rs1, vti.RegClass:$rs2,
                  GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
 
   // The choice of VFMADD here is arbitrary, vfmadd.vf and vfmacc.vf are equally
   // commutable.
+  // Fixed length vector splats
   def : Pat<(vti.Vector (riscv_fma_vl (SplatFPOp vti.ScalarRegClass:$rs1),
                                        vti.RegClass:$rd, vti.RegClass:$rs2,
                                        (vti.Mask true_mask),
                                        VLOpFrag)),
             (!cast<Instruction>("PseudoVFMADD_V" # vti.ScalarSuffix # "_" # suffix)
                  vti.RegClass:$rd, vti.ScalarRegClass:$rs1, vti.RegClass:$rs2,
                  GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
   def : Pat<(vti.Vector (riscv_fma_vl (SplatFPOp vti.ScalarRegClass:$rs1),
+                                       vti.RegClass:$rd, vti.RegClass:$rs2,
+                                       (vti.Mask V0),
+                                       VLOpFrag)),
+            (!cast<Instruction>("PseudoVFMADD_V" # vti.ScalarSuffix # "_" # suffix # "_MASK")
+                 vti.RegClass:$rd, vti.ScalarRegClass:$rs1, vti.RegClass:$rs2,
+                 (vti.Mask V0), GPR:$vl, vti.Log2SEW)>;
+  // Scalable length vector splats
+  def : Pat<(vti.Vector (riscv_fma_vl (splat_vector vti.ScalarRegClass:$rs1),
+                                       vti.RegClass:$rd, vti.RegClass:$rs2,
+                                       (vti.Mask true_mask),
+                                       VLOpFrag)),
+            (!cast<Instruction>("PseudoVFMADD_V" # vti.ScalarSuffix # "_" # suffix)
+                 vti.RegClass:$rd, vti.ScalarRegClass:$rs1, vti.RegClass:$rs2,
+                 GPR:$vl, vti.Log2SEW, TAIL_AGNOSTIC)>;
+  def : Pat<(vti.Vector (riscv_fma_vl (splat_vector vti.ScalarRegClass:$rs1),
+                                       vti.RegClass:$rd, vti.RegClass:$rs2,
+                                       (vti.Mask V0),
+                                       VLOpFrag)),
+            (!cast<Instruction>("PseudoVFMADD_V" # vti.ScalarSuffix # "_" # suffix # "_MASK")
+                 vti.RegClass:$rd, vti.ScalarRegClass:$rs1, vti.RegClass:$rs2,
+                 (vti.Mask V0), GPR:$vl, vti.Log2SEW)>;
+
+  def : Pat<(vti.Vector (riscv_fma_vl (SplatFPOp vti.ScalarRegClass:$rs1),
                                        vti.RegClass:$rd,
                                        (riscv_fneg_vl vti.RegClass:$rs2,
                                                       (vti.Mask true_mask),
                                                       VLOpFrag),
                                        (vti.Mask true_mask),
                                        VLOpFrag)),
             (!cast<Instruction>("PseudoVFMSUB_V" # vti.ScalarSuffix # "_" # suffix)
                  vti.RegClass:$rd, vti.ScalarRegClass:$rs1, vti.RegClass:$rs2,

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vfma-vp.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=riscv32 -mattr=+d,+zfh,+v -target-abi=ilp32d -riscv-v-vector-bits-min=128 \
+; RUN:   -verify-machineinstrs < %s | FileCheck %s
+; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+v -target-abi=lp64d -riscv-v-vector-bits-min=128 \
+; RUN:   -verify-machineinstrs < %s | FileCheck %s
+
+declare <2 x half> @llvm.vp.fma.v2f16(<2 x half>, <2 x half>, <2 x half>, <2 x i1>, i32)
+
+define <2 x half> @vfma_vv_v2f16(<2 x half> %va, <2 x half> %b, <2 x half> %c, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <2 x half> @llvm.vp.fma.v2f16(<2 x half> %va, <2 x half> %b, <2 x half> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x half> %v
+}
+
+define <2 x half> @vfma_vv_v2f16_unmasked(<2 x half> %va, <2 x half> %b, <2 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x half> @llvm.vp.fma.v2f16(<2 x half> %va, <2 x half> %b, <2 x half> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x half> %v
+}
+
+define <2 x half> @vfma_vf_v2f16(<2 x half> %va, half %b, <2 x half> %vc, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x half> poison, half %b, i32 0
+  %vb = shufflevector <2 x half> %elt.head, <2 x half> poison, <2 x i32> zeroinitializer
+  %v = call <2 x half> @llvm.vp.fma.v2f16(<2 x half> %va, <2 x half> %vb, <2 x half> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x half> %v
+}
+
+define <2 x half> @vfma_vf_v2f16_unmasked(<2 x half> %va, half %b, <2 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x half> poison, half %b, i32 0
+  %vb = shufflevector <2 x half> %elt.head, <2 x half> poison, <2 x i32> zeroinitializer
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x half> @llvm.vp.fma.v2f16(<2 x half> %va, <2 x half> %vb, <2 x half> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x half> %v
+}
+
+declare <4 x half> @llvm.vp.fma.v4f16(<4 x half>, <4 x half>, <4 x half>, <4 x i1>, i32)
+
+define <4 x half> @vfma_vv_v4f16(<4 x half> %va, <4 x half> %b, <4 x half> %c, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <4 x half> @llvm.vp.fma.v4f16(<4 x half> %va, <4 x half> %b, <4 x half> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x half> %v
+}
+
+define <4 x half> @vfma_vv_v4f16_unmasked(<4 x half> %va, <4 x half> %b, <4 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x half> @llvm.vp.fma.v4f16(<4 x half> %va, <4 x half> %b, <4 x half> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x half> %v
+}
+
+define <4 x half> @vfma_vf_v4f16(<4 x half> %va, half %b, <4 x half> %vc, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x half> poison, half %b, i32 0
+  %vb = shufflevector <4 x half> %elt.head, <4 x half> poison, <4 x i32> zeroinitializer
+  %v = call <4 x half> @llvm.vp.fma.v4f16(<4 x half> %va, <4 x half> %vb, <4 x half> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x half> %v
+}
+
+define <4 x half> @vfma_vf_v4f16_unmasked(<4 x half> %va, half %b, <4 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x half> poison, half %b, i32 0
+  %vb = shufflevector <4 x half> %elt.head, <4 x half> poison, <4 x i32> zeroinitializer
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x half> @llvm.vp.fma.v4f16(<4 x half> %va, <4 x half> %vb, <4 x half> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x half> %v
+}
+
+declare <8 x half> @llvm.vp.fma.v8f16(<8 x half>, <8 x half>, <8 x half>, <8 x i1>, i32)
+
+define <8 x half> @vfma_vv_v8f16(<8 x half> %va, <8 x half> %b, <8 x half> %c, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <8 x half> @llvm.vp.fma.v8f16(<8 x half> %va, <8 x half> %b, <8 x half> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x half> %v
+}
+
+define <8 x half> @vfma_vv_v8f16_unmasked(<8 x half> %va, <8 x half> %b, <8 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x half> @llvm.vp.fma.v8f16(<8 x half> %va, <8 x half> %b, <8 x half> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x half> %v
+}
+
+define <8 x half> @vfma_vf_v8f16(<8 x half> %va, half %b, <8 x half> %vc, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x half> poison, half %b, i32 0
+  %vb = shufflevector <8 x half> %elt.head, <8 x half> poison, <8 x i32> zeroinitializer
+  %v = call <8 x half> @llvm.vp.fma.v8f16(<8 x half> %va, <8 x half> %vb, <8 x half> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x half> %v
+}
+
+define <8 x half> @vfma_vf_v8f16_unmasked(<8 x half> %va, half %b, <8 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x half> poison, half %b, i32 0
+  %vb = shufflevector <8 x half> %elt.head, <8 x half> poison, <8 x i32> zeroinitializer
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x half> @llvm.vp.fma.v8f16(<8 x half> %va, <8 x half> %vb, <8 x half> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x half> %v
+}
+
+declare <16 x half> @llvm.vp.fma.v16f16(<16 x half>, <16 x half>, <16 x half>, <16 x i1>, i32)
+
+define <16 x half> @vfma_vv_v16f16(<16 x half> %va, <16 x half> %b, <16 x half> %c, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <16 x half> @llvm.vp.fma.v16f16(<16 x half> %va, <16 x half> %b, <16 x half> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x half> %v
+}
+
+define <16 x half> @vfma_vv_v16f16_unmasked(<16 x half> %va, <16 x half> %b, <16 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x half> @llvm.vp.fma.v16f16(<16 x half> %va, <16 x half> %b, <16 x half> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x half> %v
+}
+
+define <16 x half> @vfma_vf_v16f16(<16 x half> %va, half %b, <16 x half> %vc, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x half> poison, half %b, i32 0
+  %vb = shufflevector <16 x half> %elt.head, <16 x half> poison, <16 x i32> zeroinitializer
+  %v = call <16 x half> @llvm.vp.fma.v16f16(<16 x half> %va, <16 x half> %vb, <16 x half> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x half> %v
+}
+
+define <16 x half> @vfma_vf_v16f16_unmasked(<16 x half> %va, half %b, <16 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x half> poison, half %b, i32 0
+  %vb = shufflevector <16 x half> %elt.head, <16 x half> poison, <16 x i32> zeroinitializer
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x half> @llvm.vp.fma.v16f16(<16 x half> %va, <16 x half> %vb, <16 x half> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x half> %v
+}
+
+declare <2 x float> @llvm.vp.fma.v2f32(<2 x float>, <2 x float>, <2 x float>, <2 x i1>, i32)
+
+define <2 x float> @vfma_vv_v2f32(<2 x float> %va, <2 x float> %b, <2 x float> %c, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <2 x float> @llvm.vp.fma.v2f32(<2 x float> %va, <2 x float> %b, <2 x float> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x float> %v
+}
+
+define <2 x float> @vfma_vv_v2f32_unmasked(<2 x float> %va, <2 x float> %b, <2 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x float> @llvm.vp.fma.v2f32(<2 x float> %va, <2 x float> %b, <2 x float> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x float> %v
+}
+
+define <2 x float> @vfma_vf_v2f32(<2 x float> %va, float %b, <2 x float> %vc, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x float> poison, float %b, i32 0
+  %vb = shufflevector <2 x float> %elt.head, <2 x float> poison, <2 x i32> zeroinitializer
+  %v = call <2 x float> @llvm.vp.fma.v2f32(<2 x float> %va, <2 x float> %vb, <2 x float> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x float> %v
+}
+
+define <2 x float> @vfma_vf_v2f32_unmasked(<2 x float> %va, float %b, <2 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x float> poison, float %b, i32 0
+  %vb = shufflevector <2 x float> %elt.head, <2 x float> poison, <2 x i32> zeroinitializer
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x float> @llvm.vp.fma.v2f32(<2 x float> %va, <2 x float> %vb, <2 x float> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x float> %v
+}
+
+declare <4 x float> @llvm.vp.fma.v4f32(<4 x float>, <4 x float>, <4 x float>, <4 x i1>, i32)
+
+define <4 x float> @vfma_vv_v4f32(<4 x float> %va, <4 x float> %b, <4 x float> %c, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <4 x float> @llvm.vp.fma.v4f32(<4 x float> %va, <4 x float> %b, <4 x float> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x float> %v
+}
+
+define <4 x float> @vfma_vv_v4f32_unmasked(<4 x float> %va, <4 x float> %b, <4 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x float> @llvm.vp.fma.v4f32(<4 x float> %va, <4 x float> %b, <4 x float> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x float> %v
+}
+
+define <4 x float> @vfma_vf_v4f32(<4 x float> %va, float %b, <4 x float> %vc, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x float> poison, float %b, i32 0
+  %vb = shufflevector <4 x float> %elt.head, <4 x float> poison, <4 x i32> zeroinitializer
+  %v = call <4 x float> @llvm.vp.fma.v4f32(<4 x float> %va, <4 x float> %vb, <4 x float> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x float> %v
+}
+
+define <4 x float> @vfma_vf_v4f32_unmasked(<4 x float> %va, float %b, <4 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x float> poison, float %b, i32 0
+  %vb = shufflevector <4 x float> %elt.head, <4 x float> poison, <4 x i32> zeroinitializer
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x float> @llvm.vp.fma.v4f32(<4 x float> %va, <4 x float> %vb, <4 x float> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x float> %v
+}
+
+declare <8 x float> @llvm.vp.fma.v8f32(<8 x float>, <8 x float>, <8 x float>, <8 x i1>, i32)
+
+define <8 x float> @vfma_vv_v8f32(<8 x float> %va, <8 x float> %b, <8 x float> %c, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <8 x float> @llvm.vp.fma.v8f32(<8 x float> %va, <8 x float> %b, <8 x float> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x float> %v
+}
+
+define <8 x float> @vfma_vv_v8f32_unmasked(<8 x float> %va, <8 x float> %b, <8 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x float> @llvm.vp.fma.v8f32(<8 x float> %va, <8 x float> %b, <8 x float> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x float> %v
+}
+
+define <8 x float> @vfma_vf_v8f32(<8 x float> %va, float %b, <8 x float> %vc, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x float> poison, float %b, i32 0
+  %vb = shufflevector <8 x float> %elt.head, <8 x float> poison, <8 x i32> zeroinitializer
+  %v = call <8 x float> @llvm.vp.fma.v8f32(<8 x float> %va, <8 x float> %vb, <8 x float> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x float> %v
+}
+
+define <8 x float> @vfma_vf_v8f32_unmasked(<8 x float> %va, float %b, <8 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x float> poison, float %b, i32 0
+  %vb = shufflevector <8 x float> %elt.head, <8 x float> poison, <8 x i32> zeroinitializer
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x float> @llvm.vp.fma.v8f32(<8 x float> %va, <8 x float> %vb, <8 x float> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x float> %v
+}
+
+declare <16 x float> @llvm.vp.fma.v16f32(<16 x float>, <16 x float>, <16 x float>, <16 x i1>, i32)
+
+define <16 x float> @vfma_vv_v16f32(<16 x float> %va, <16 x float> %b, <16 x float> %c, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v12, v8, v16, v0.t
+; CHECK-NEXT:    vmv4r.v v8, v12
+; CHECK-NEXT:    ret
+  %v = call <16 x float> @llvm.vp.fma.v16f32(<16 x float> %va, <16 x float> %b, <16 x float> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x float> %v
+}
+
+define <16 x float> @vfma_vv_v16f32_unmasked(<16 x float> %va, <16 x float> %b, <16 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v12, v16
+; CHECK-NEXT:    ret
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x float> @llvm.vp.fma.v16f32(<16 x float> %va, <16 x float> %b, <16 x float> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x float> %v
+}
+
+define <16 x float> @vfma_vf_v16f32(<16 x float> %va, float %b, <16 x float> %vc, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x float> poison, float %b, i32 0
+  %vb = shufflevector <16 x float> %elt.head, <16 x float> poison, <16 x i32> zeroinitializer
+  %v = call <16 x float> @llvm.vp.fma.v16f32(<16 x float> %va, <16 x float> %vb, <16 x float> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x float> %v
+}
+
+define <16 x float> @vfma_vf_v16f32_unmasked(<16 x float> %va, float %b, <16 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x float> poison, float %b, i32 0
+  %vb = shufflevector <16 x float> %elt.head, <16 x float> poison, <16 x i32> zeroinitializer
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x float> @llvm.vp.fma.v16f32(<16 x float> %va, <16 x float> %vb, <16 x float> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x float> %v
+}
+
+declare <2 x double> @llvm.vp.fma.v2f64(<2 x double>, <2 x double>, <2 x double>, <2 x i1>, i32)
+
+define <2 x double> @vfma_vv_v2f64(<2 x double> %va, <2 x double> %b, <2 x double> %c, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <2 x double> @llvm.vp.fma.v2f64(<2 x double> %va, <2 x double> %b, <2 x double> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x double> %v
+}
+
+define <2 x double> @vfma_vv_v2f64_unmasked(<2 x double> %va, <2 x double> %b, <2 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v2f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x double> @llvm.vp.fma.v2f64(<2 x double> %va, <2 x double> %b, <2 x double> %c, <2 x i1> %m, i32 %evl)
+  ret <2 x double> %v
+}
+
+define <2 x double> @vfma_vf_v2f64(<2 x double> %va, double %b, <2 x double> %vc, <2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x double> poison, double %b, i32 0
+  %vb = shufflevector <2 x double> %elt.head, <2 x double> poison, <2 x i32> zeroinitializer
+  %v = call <2 x double> @llvm.vp.fma.v2f64(<2 x double> %va, <2 x double> %vb, <2 x double> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x double> %v
+}
+
+define <2 x double> @vfma_vf_v2f64_unmasked(<2 x double> %va, double %b, <2 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v2f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <2 x double> poison, double %b, i32 0
+  %vb = shufflevector <2 x double> %elt.head, <2 x double> poison, <2 x i32> zeroinitializer
+  %head = insertelement <2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <2 x i1> %head, <2 x i1> poison, <2 x i32> zeroinitializer
+  %v = call <2 x double> @llvm.vp.fma.v2f64(<2 x double> %va, <2 x double> %vb, <2 x double> %vc, <2 x i1> %m, i32 %evl)
+  ret <2 x double> %v
+}
+
+declare <4 x double> @llvm.vp.fma.v4f64(<4 x double>, <4 x double>, <4 x double>, <4 x i1>, i32)
+
+define <4 x double> @vfma_vv_v4f64(<4 x double> %va, <4 x double> %b, <4 x double> %c, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <4 x double> @llvm.vp.fma.v4f64(<4 x double> %va, <4 x double> %b, <4 x double> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x double> %v
+}
+
+define <4 x double> @vfma_vv_v4f64_unmasked(<4 x double> %va, <4 x double> %b, <4 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v4f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x double> @llvm.vp.fma.v4f64(<4 x double> %va, <4 x double> %b, <4 x double> %c, <4 x i1> %m, i32 %evl)
+  ret <4 x double> %v
+}
+
+define <4 x double> @vfma_vf_v4f64(<4 x double> %va, double %b, <4 x double> %vc, <4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x double> poison, double %b, i32 0
+  %vb = shufflevector <4 x double> %elt.head, <4 x double> poison, <4 x i32> zeroinitializer
+  %v = call <4 x double> @llvm.vp.fma.v4f64(<4 x double> %va, <4 x double> %vb, <4 x double> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x double> %v
+}
+
+define <4 x double> @vfma_vf_v4f64_unmasked(<4 x double> %va, double %b, <4 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v4f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <4 x double> poison, double %b, i32 0
+  %vb = shufflevector <4 x double> %elt.head, <4 x double> poison, <4 x i32> zeroinitializer
+  %head = insertelement <4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <4 x i1> %head, <4 x i1> poison, <4 x i32> zeroinitializer
+  %v = call <4 x double> @llvm.vp.fma.v4f64(<4 x double> %va, <4 x double> %vb, <4 x double> %vc, <4 x i1> %m, i32 %evl)
+  ret <4 x double> %v
+}
+
+declare <8 x double> @llvm.vp.fma.v8f64(<8 x double>, <8 x double>, <8 x double>, <8 x i1>, i32)
+
+define <8 x double> @vfma_vv_v8f64(<8 x double> %va, <8 x double> %b, <8 x double> %c, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v12, v8, v16, v0.t
+; CHECK-NEXT:    vmv4r.v v8, v12
+; CHECK-NEXT:    ret
+  %v = call <8 x double> @llvm.vp.fma.v8f64(<8 x double> %va, <8 x double> %b, <8 x double> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x double> %v
+}
+
+define <8 x double> @vfma_vv_v8f64_unmasked(<8 x double> %va, <8 x double> %b, <8 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v8f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v12, v16
+; CHECK-NEXT:    ret
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x double> @llvm.vp.fma.v8f64(<8 x double> %va, <8 x double> %b, <8 x double> %c, <8 x i1> %m, i32 %evl)
+  ret <8 x double> %v
+}
+
+define <8 x double> @vfma_vf_v8f64(<8 x double> %va, double %b, <8 x double> %vc, <8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x double> poison, double %b, i32 0
+  %vb = shufflevector <8 x double> %elt.head, <8 x double> poison, <8 x i32> zeroinitializer
+  %v = call <8 x double> @llvm.vp.fma.v8f64(<8 x double> %va, <8 x double> %vb, <8 x double> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x double> %v
+}
+
+define <8 x double> @vfma_vf_v8f64_unmasked(<8 x double> %va, double %b, <8 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v8f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <8 x double> poison, double %b, i32 0
+  %vb = shufflevector <8 x double> %elt.head, <8 x double> poison, <8 x i32> zeroinitializer
+  %head = insertelement <8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <8 x i1> %head, <8 x i1> poison, <8 x i32> zeroinitializer
+  %v = call <8 x double> @llvm.vp.fma.v8f64(<8 x double> %va, <8 x double> %vb, <8 x double> %vc, <8 x i1> %m, i32 %evl)
+  ret <8 x double> %v
+}
+
+declare <16 x double> @llvm.vp.fma.v16f64(<16 x double>, <16 x double>, <16 x double>, <16 x i1>, i32)
+
+define <16 x double> @vfma_vv_v16f64(<16 x double> %va, <16 x double> %b, <16 x double> %c, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, mu
+; CHECK-NEXT:    vle64.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vv v16, v8, v24, v0.t
+; CHECK-NEXT:    vmv8r.v v8, v16
+; CHECK-NEXT:    ret
+  %v = call <16 x double> @llvm.vp.fma.v16f64(<16 x double> %va, <16 x double> %b, <16 x double> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x double> %v
+}
+
+define <16 x double> @vfma_vv_v16f64_unmasked(<16 x double> %va, <16 x double> %b, <16 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_v16f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, mu
+; CHECK-NEXT:    vle64.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v16, v24
+; CHECK-NEXT:    ret
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x double> @llvm.vp.fma.v16f64(<16 x double> %va, <16 x double> %b, <16 x double> %c, <16 x i1> %m, i32 %evl)
+  ret <16 x double> %v
+}
+
+define <16 x double> @vfma_vf_v16f64(<16 x double> %va, double %b, <16 x double> %vc, <16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x double> poison, double %b, i32 0
+  %vb = shufflevector <16 x double> %elt.head, <16 x double> poison, <16 x i32> zeroinitializer
+  %v = call <16 x double> @llvm.vp.fma.v16f64(<16 x double> %va, <16 x double> %vb, <16 x double> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x double> %v
+}
+
+define <16 x double> @vfma_vf_v16f64_unmasked(<16 x double> %va, double %b, <16 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_v16f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <16 x double> poison, double %b, i32 0
+  %vb = shufflevector <16 x double> %elt.head, <16 x double> poison, <16 x i32> zeroinitializer
+  %head = insertelement <16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <16 x i1> %head, <16 x i1> poison, <16 x i32> zeroinitializer
+  %v = call <16 x double> @llvm.vp.fma.v16f64(<16 x double> %va, <16 x double> %vb, <16 x double> %vc, <16 x i1> %m, i32 %evl)
+  ret <16 x double> %v
+}

llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=riscv32 -mattr=+d,+zfh,+v -target-abi=ilp32d \
+; RUN:     -verify-machineinstrs < %s | FileCheck %s
+; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+v -target-abi=lp64d \
+; RUN:     -verify-machineinstrs < %s | FileCheck %s
+
+declare <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half>, <vscale x 1 x half>, <vscale x 1 x half>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x half> @vfma_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %b, <vscale x 1 x half> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %b, <vscale x 1 x half> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vfma_vv_nxv1f16_unmasked(<vscale x 1 x half> %va, <vscale x 1 x half> %b, <vscale x 1 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %b, <vscale x 1 x half> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vfma_vf_nxv1f16(<vscale x 1 x half> %va, half %b, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 1 x half> %elt.head, <vscale x 1 x half> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vfma_vf_nxv1f16_commute(<vscale x 1 x half> %va, half %b, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 1 x half> %elt.head, <vscale x 1 x half> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %vb, <vscale x 1 x half> %va, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vfma_vf_nxv1f16_unmasked(<vscale x 1 x half> %va, half %b, <vscale x 1 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 1 x half> %elt.head, <vscale x 1 x half> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+define <vscale x 1 x half> @vfma_vf_nxv1f16_unmasked_commute(<vscale x 1 x half> %va, half %b, <vscale x 1 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 1 x half> %elt.head, <vscale x 1 x half> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x half> @llvm.vp.fma.nxv1f16(<vscale x 1 x half> %vb, <vscale x 1 x half> %va, <vscale x 1 x half> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x half> %v
+}
+
+declare <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half>, <vscale x 2 x half>, <vscale x 2 x half>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x half> @vfma_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %b, <vscale x 2 x half> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %b, <vscale x 2 x half> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vfma_vv_nxv2f16_unmasked(<vscale x 2 x half> %va, <vscale x 2 x half> %b, <vscale x 2 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %b, <vscale x 2 x half> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vfma_vf_nxv2f16(<vscale x 2 x half> %va, half %b, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 2 x half> %elt.head, <vscale x 2 x half> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vfma_vf_nxv2f16_commute(<vscale x 2 x half> %va, half %b, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 2 x half> %elt.head, <vscale x 2 x half> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %vb, <vscale x 2 x half> %va, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vfma_vf_nxv2f16_unmasked(<vscale x 2 x half> %va, half %b, <vscale x 2 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 2 x half> %elt.head, <vscale x 2 x half> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+define <vscale x 2 x half> @vfma_vf_nxv2f16_unmasked_commute(<vscale x 2 x half> %va, half %b, <vscale x 2 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 2 x half> %elt.head, <vscale x 2 x half> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x half> @llvm.vp.fma.nxv2f16(<vscale x 2 x half> %vb, <vscale x 2 x half> %va, <vscale x 2 x half> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x half> %v
+}
+
+declare <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half>, <vscale x 4 x half>, <vscale x 4 x half>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x half> @vfma_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %b, <vscale x 4 x half> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %b, <vscale x 4 x half> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vfma_vv_nxv4f16_unmasked(<vscale x 4 x half> %va, <vscale x 4 x half> %b, <vscale x 4 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %b, <vscale x 4 x half> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vfma_vf_nxv4f16(<vscale x 4 x half> %va, half %b, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 4 x half> %elt.head, <vscale x 4 x half> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vfma_vf_nxv4f16_commute(<vscale x 4 x half> %va, half %b, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 4 x half> %elt.head, <vscale x 4 x half> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %vb, <vscale x 4 x half> %va, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vfma_vf_nxv4f16_unmasked(<vscale x 4 x half> %va, half %b, <vscale x 4 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 4 x half> %elt.head, <vscale x 4 x half> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+define <vscale x 4 x half> @vfma_vf_nxv4f16_unmasked_commute(<vscale x 4 x half> %va, half %b, <vscale x 4 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 4 x half> %elt.head, <vscale x 4 x half> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x half> @llvm.vp.fma.nxv4f16(<vscale x 4 x half> %vb, <vscale x 4 x half> %va, <vscale x 4 x half> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x half> %v
+}
+
+declare <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x half>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x half> @vfma_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %b, <vscale x 8 x half> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %b, <vscale x 8 x half> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vfma_vv_nxv8f16_unmasked(<vscale x 8 x half> %va, <vscale x 8 x half> %b, <vscale x 8 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %b, <vscale x 8 x half> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vfma_vf_nxv8f16(<vscale x 8 x half> %va, half %b, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 8 x half> %elt.head, <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vfma_vf_nxv8f16_commute(<vscale x 8 x half> %va, half %b, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 8 x half> %elt.head, <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %vb, <vscale x 8 x half> %va, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vfma_vf_nxv8f16_unmasked(<vscale x 8 x half> %va, half %b, <vscale x 8 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 8 x half> %elt.head, <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+define <vscale x 8 x half> @vfma_vf_nxv8f16_unmasked_commute(<vscale x 8 x half> %va, half %b, <vscale x 8 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 8 x half> %elt.head, <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x half> @llvm.vp.fma.nxv8f16(<vscale x 8 x half> %vb, <vscale x 8 x half> %va, <vscale x 8 x half> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x half> %v
+}
+
+declare <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half>, <vscale x 16 x half>, <vscale x 16 x half>, <vscale x 16 x i1>, i32)
+
+define <vscale x 16 x half> @vfma_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %b, <vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv16f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v12, v8, v16, v0.t
+; CHECK-NEXT:    vmv4r.v v8, v12
+; CHECK-NEXT:    ret
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %b, <vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vfma_vv_nxv16f16_unmasked(<vscale x 16 x half> %va, <vscale x 16 x half> %b, <vscale x 16 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv16f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v12, v16
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %b, <vscale x 16 x half> %c, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vfma_vf_nxv16f16(<vscale x 16 x half> %va, half %b, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vfma_vf_nxv16f16_commute(<vscale x 16 x half> %va, half %b, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %vb, <vscale x 16 x half> %va, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vfma_vf_nxv16f16_unmasked(<vscale x 16 x half> %va, half %b, <vscale x 16 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+define <vscale x 16 x half> @vfma_vf_nxv16f16_unmasked_commute(<vscale x 16 x half> %va, half %b, <vscale x 16 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 16 x half> %elt.head, <vscale x 16 x half> poison, <vscale x 16 x i32> zeroinitializer
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x half> @llvm.vp.fma.nxv16f16(<vscale x 16 x half> %vb, <vscale x 16 x half> %va, <vscale x 16 x half> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x half> %v
+}
+
+declare <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half>, <vscale x 32 x half>, <vscale x 32 x half>, <vscale x 32 x i1>, i32)
+
+define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %b, <vscale x 32 x half> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv32f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re16.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e16, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vv v16, v8, v24, v0.t
+; CHECK-NEXT:    vmv8r.v v8, v16
+; CHECK-NEXT:    ret
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %b, <vscale x 32 x half> %c, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vfma_vv_nxv32f16_unmasked(<vscale x 32 x half> %va, <vscale x 32 x half> %b, <vscale x 32 x half> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv32f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re16.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e16, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v16, v24
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 32 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 32 x i1> %head, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %b, <vscale x 32 x half> %c, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vfma_vf_nxv32f16(<vscale x 32 x half> %va, half %b, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv32f16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vfma_vf_nxv32f16_commute(<vscale x 32 x half> %va, half %b, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv32f16_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %vb, <vscale x 32 x half> %va, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vfma_vf_nxv32f16_unmasked(<vscale x 32 x half> %va, half %b, <vscale x 32 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv32f16_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
+  %head = insertelement <vscale x 32 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 32 x i1> %head, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x half> %vb, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+define <vscale x 32 x half> @vfma_vf_nxv32f16_unmasked_commute(<vscale x 32 x half> %va, half %b, <vscale x 32 x half> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv32f16_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e16, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 32 x half> poison, half %b, i32 0
+  %vb = shufflevector <vscale x 32 x half> %elt.head, <vscale x 32 x half> poison, <vscale x 32 x i32> zeroinitializer
+  %head = insertelement <vscale x 32 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 32 x i1> %head, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+  %v = call <vscale x 32 x half> @llvm.vp.fma.nxv32f16(<vscale x 32 x half> %vb, <vscale x 32 x half> %va, <vscale x 32 x half> %vc, <vscale x 32 x i1> %m, i32 %evl)
+  ret <vscale x 32 x half> %v
+}
+
+declare <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float>, <vscale x 1 x float>, <vscale x 1 x float>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x float> @vfma_vv_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %b, <vscale x 1 x float> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %b, <vscale x 1 x float> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vfma_vv_nxv1f32_unmasked(<vscale x 1 x float> %va, <vscale x 1 x float> %b, <vscale x 1 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %b, <vscale x 1 x float> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vfma_vf_nxv1f32(<vscale x 1 x float> %va, float %b, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 1 x float> %elt.head, <vscale x 1 x float> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vfma_vf_nxv1f32_commute(<vscale x 1 x float> %va, float %b, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f32_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 1 x float> %elt.head, <vscale x 1 x float> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %vb, <vscale x 1 x float> %va, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vfma_vf_nxv1f32_unmasked(<vscale x 1 x float> %va, float %b, <vscale x 1 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 1 x float> %elt.head, <vscale x 1 x float> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+define <vscale x 1 x float> @vfma_vf_nxv1f32_unmasked_commute(<vscale x 1 x float> %va, float %b, <vscale x 1 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f32_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 1 x float> %elt.head, <vscale x 1 x float> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x float> @llvm.vp.fma.nxv1f32(<vscale x 1 x float> %vb, <vscale x 1 x float> %va, <vscale x 1 x float> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x float> %v
+}
+
+declare <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float>, <vscale x 2 x float>, <vscale x 2 x float>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x float> @vfma_vv_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %b, <vscale x 2 x float> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %b, <vscale x 2 x float> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vfma_vv_nxv2f32_unmasked(<vscale x 2 x float> %va, <vscale x 2 x float> %b, <vscale x 2 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %b, <vscale x 2 x float> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vfma_vf_nxv2f32(<vscale x 2 x float> %va, float %b, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 2 x float> %elt.head, <vscale x 2 x float> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vfma_vf_nxv2f32_commute(<vscale x 2 x float> %va, float %b, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f32_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 2 x float> %elt.head, <vscale x 2 x float> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %vb, <vscale x 2 x float> %va, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vfma_vf_nxv2f32_unmasked(<vscale x 2 x float> %va, float %b, <vscale x 2 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 2 x float> %elt.head, <vscale x 2 x float> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+define <vscale x 2 x float> @vfma_vf_nxv2f32_unmasked_commute(<vscale x 2 x float> %va, float %b, <vscale x 2 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f32_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 2 x float> %elt.head, <vscale x 2 x float> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x float> @llvm.vp.fma.nxv2f32(<vscale x 2 x float> %vb, <vscale x 2 x float> %va, <vscale x 2 x float> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x float> %v
+}
+
+declare <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x float>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x float> @vfma_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %b, <vscale x 4 x float> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %b, <vscale x 4 x float> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vfma_vv_nxv4f32_unmasked(<vscale x 4 x float> %va, <vscale x 4 x float> %b, <vscale x 4 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %b, <vscale x 4 x float> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vfma_vf_nxv4f32(<vscale x 4 x float> %va, float %b, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 4 x float> %elt.head, <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vfma_vf_nxv4f32_commute(<vscale x 4 x float> %va, float %b, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f32_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 4 x float> %elt.head, <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %vb, <vscale x 4 x float> %va, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vfma_vf_nxv4f32_unmasked(<vscale x 4 x float> %va, float %b, <vscale x 4 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 4 x float> %elt.head, <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+define <vscale x 4 x float> @vfma_vf_nxv4f32_unmasked_commute(<vscale x 4 x float> %va, float %b, <vscale x 4 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f32_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 4 x float> %elt.head, <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x float> @llvm.vp.fma.nxv4f32(<vscale x 4 x float> %vb, <vscale x 4 x float> %va, <vscale x 4 x float> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x float> %v
+}
+
+declare <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float>, <vscale x 8 x float>, <vscale x 8 x float>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x float> @vfma_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %b, <vscale x 8 x float> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v12, v8, v16, v0.t
+; CHECK-NEXT:    vmv4r.v v8, v12
+; CHECK-NEXT:    ret
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %b, <vscale x 8 x float> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vfma_vv_nxv8f32_unmasked(<vscale x 8 x float> %va, <vscale x 8 x float> %b, <vscale x 8 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v12, v16
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %b, <vscale x 8 x float> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vfma_vf_nxv8f32(<vscale x 8 x float> %va, float %b, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 8 x float> %elt.head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vfma_vf_nxv8f32_commute(<vscale x 8 x float> %va, float %b, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f32_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 8 x float> %elt.head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %vb, <vscale x 8 x float> %va, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vfma_vf_nxv8f32_unmasked(<vscale x 8 x float> %va, float %b, <vscale x 8 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 8 x float> %elt.head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+define <vscale x 8 x float> @vfma_vf_nxv8f32_unmasked_commute(<vscale x 8 x float> %va, float %b, <vscale x 8 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f32_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 8 x float> %elt.head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x float> @llvm.vp.fma.nxv8f32(<vscale x 8 x float> %vb, <vscale x 8 x float> %va, <vscale x 8 x float> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x float> %v
+}
+
+declare <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float>, <vscale x 16 x float>, <vscale x 16 x float>, <vscale x 16 x i1>, i32)
+
+define <vscale x 16 x float> @vfma_vv_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %b, <vscale x 16 x float> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv16f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re32.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e32, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vv v16, v8, v24, v0.t
+; CHECK-NEXT:    vmv8r.v v8, v16
+; CHECK-NEXT:    ret
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %b, <vscale x 16 x float> %c, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vfma_vv_nxv16f32_unmasked(<vscale x 16 x float> %va, <vscale x 16 x float> %b, <vscale x 16 x float> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv16f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re32.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e32, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v16, v24
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %b, <vscale x 16 x float> %c, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vfma_vf_nxv16f32(<vscale x 16 x float> %va, float %b, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f32:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 16 x float> %elt.head, <vscale x 16 x float> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vfma_vf_nxv16f32_commute(<vscale x 16 x float> %va, float %b, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f32_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 16 x float> %elt.head, <vscale x 16 x float> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %vb, <vscale x 16 x float> %va, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vfma_vf_nxv16f32_unmasked(<vscale x 16 x float> %va, float %b, <vscale x 16 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f32_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 16 x float> %elt.head, <vscale x 16 x float> poison, <vscale x 16 x i32> zeroinitializer
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x float> %vb, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+define <vscale x 16 x float> @vfma_vf_nxv16f32_unmasked_commute(<vscale x 16 x float> %va, float %b, <vscale x 16 x float> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv16f32_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 16 x float> poison, float %b, i32 0
+  %vb = shufflevector <vscale x 16 x float> %elt.head, <vscale x 16 x float> poison, <vscale x 16 x i32> zeroinitializer
+  %head = insertelement <vscale x 16 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 16 x i1> %head, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+  %v = call <vscale x 16 x float> @llvm.vp.fma.nxv16f32(<vscale x 16 x float> %vb, <vscale x 16 x float> %va, <vscale x 16 x float> %vc, <vscale x 16 x i1> %m, i32 %evl)
+  ret <vscale x 16 x float> %v
+}
+
+declare <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double>, <vscale x 1 x double>, <vscale x 1 x double>, <vscale x 1 x i1>, i32)
+
+define <vscale x 1 x double> @vfma_vv_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %b, <vscale x 1 x double> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vv v9, v8, v10, v0.t
+; CHECK-NEXT:    vmv1r.v v8, v9
+; CHECK-NEXT:    ret
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %b, <vscale x 1 x double> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vfma_vv_nxv1f64_unmasked(<vscale x 1 x double> %va, <vscale x 1 x double> %b, <vscale x 1 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv1f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v9, v10
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %b, <vscale x 1 x double> %c, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vfma_vf_nxv1f64(<vscale x 1 x double> %va, double %b, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 1 x double> %elt.head, <vscale x 1 x double> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vfma_vf_nxv1f64_commute(<vscale x 1 x double> %va, double %b, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f64_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 1 x double> %elt.head, <vscale x 1 x double> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %vb, <vscale x 1 x double> %va, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vfma_vf_nxv1f64_unmasked(<vscale x 1 x double> %va, double %b, <vscale x 1 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 1 x double> %elt.head, <vscale x 1 x double> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+define <vscale x 1 x double> @vfma_vf_nxv1f64_unmasked_commute(<vscale x 1 x double> %va, double %b, <vscale x 1 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv1f64_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v9
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 1 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 1 x double> %elt.head, <vscale x 1 x double> poison, <vscale x 1 x i32> zeroinitializer
+  %head = insertelement <vscale x 1 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 1 x i1> %head, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+  %v = call <vscale x 1 x double> @llvm.vp.fma.nxv1f64(<vscale x 1 x double> %vb, <vscale x 1 x double> %va, <vscale x 1 x double> %vc, <vscale x 1 x i1> %m, i32 %evl)
+  ret <vscale x 1 x double> %v
+}
+
+declare <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x double>, <vscale x 2 x i1>, i32)
+
+define <vscale x 2 x double> @vfma_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %b, <vscale x 2 x double> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vv v10, v8, v12, v0.t
+; CHECK-NEXT:    vmv2r.v v8, v10
+; CHECK-NEXT:    ret
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %b, <vscale x 2 x double> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vfma_vv_nxv2f64_unmasked(<vscale x 2 x double> %va, <vscale x 2 x double> %b, <vscale x 2 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv2f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v10, v12
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %b, <vscale x 2 x double> %c, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vfma_vf_nxv2f64(<vscale x 2 x double> %va, double %b, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 2 x double> %elt.head, <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vfma_vf_nxv2f64_commute(<vscale x 2 x double> %va, double %b, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f64_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 2 x double> %elt.head, <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %vb, <vscale x 2 x double> %va, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vfma_vf_nxv2f64_unmasked(<vscale x 2 x double> %va, double %b, <vscale x 2 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 2 x double> %elt.head, <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+define <vscale x 2 x double> @vfma_vf_nxv2f64_unmasked_commute(<vscale x 2 x double> %va, double %b, <vscale x 2 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv2f64_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v10
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 2 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 2 x double> %elt.head, <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
+  %head = insertelement <vscale x 2 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 2 x i1> %head, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+  %v = call <vscale x 2 x double> @llvm.vp.fma.nxv2f64(<vscale x 2 x double> %vb, <vscale x 2 x double> %va, <vscale x 2 x double> %vc, <vscale x 2 x i1> %m, i32 %evl)
+  ret <vscale x 2 x double> %v
+}
+
+declare <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double>, <vscale x 4 x double>, <vscale x 4 x double>, <vscale x 4 x i1>, i32)
+
+define <vscale x 4 x double> @vfma_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %b, <vscale x 4 x double> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vv v12, v8, v16, v0.t
+; CHECK-NEXT:    vmv4r.v v8, v12
+; CHECK-NEXT:    ret
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %b, <vscale x 4 x double> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vfma_vv_nxv4f64_unmasked(<vscale x 4 x double> %va, <vscale x 4 x double> %b, <vscale x 4 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv4f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v12, v16
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %b, <vscale x 4 x double> %c, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vfma_vf_nxv4f64(<vscale x 4 x double> %va, double %b, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 4 x double> %elt.head, <vscale x 4 x double> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vfma_vf_nxv4f64_commute(<vscale x 4 x double> %va, double %b, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f64_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 4 x double> %elt.head, <vscale x 4 x double> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %vb, <vscale x 4 x double> %va, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vfma_vf_nxv4f64_unmasked(<vscale x 4 x double> %va, double %b, <vscale x 4 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 4 x double> %elt.head, <vscale x 4 x double> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+define <vscale x 4 x double> @vfma_vf_nxv4f64_unmasked_commute(<vscale x 4 x double> %va, double %b, <vscale x 4 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv4f64_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v12
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 4 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 4 x double> %elt.head, <vscale x 4 x double> poison, <vscale x 4 x i32> zeroinitializer
+  %head = insertelement <vscale x 4 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 4 x i1> %head, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+  %v = call <vscale x 4 x double> @llvm.vp.fma.nxv4f64(<vscale x 4 x double> %vb, <vscale x 4 x double> %va, <vscale x 4 x double> %vc, <vscale x 4 x i1> %m, i32 %evl)
+  ret <vscale x 4 x double> %v
+}
+
+declare <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double>, <vscale x 8 x double>, <vscale x 8 x double>, <vscale x 8 x i1>, i32)
+
+define <vscale x 8 x double> @vfma_vv_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %b, <vscale x 8 x double> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re64.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vv v16, v8, v24, v0.t
+; CHECK-NEXT:    vmv8r.v v8, v16
+; CHECK-NEXT:    ret
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %b, <vscale x 8 x double> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vfma_vv_nxv8f64_unmasked(<vscale x 8 x double> %va, <vscale x 8 x double> %b, <vscale x 8 x double> %c, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vv_nxv8f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vl8re64.v v24, (a0)
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vv v8, v16, v24
+; CHECK-NEXT:    ret
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %b, <vscale x 8 x double> %c, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vfma_vf_nxv8f64(<vscale x 8 x double> %va, double %b, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f64:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 8 x double> %elt.head, <vscale x 8 x double> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vfma_vf_nxv8f64_commute(<vscale x 8 x double> %va, double %b, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f64_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, tu, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16, v0.t
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 8 x double> %elt.head, <vscale x 8 x double> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %vb, <vscale x 8 x double> %va, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vfma_vf_nxv8f64_unmasked(<vscale x 8 x double> %va, double %b, <vscale x 8 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f64_unmasked:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 8 x double> %elt.head, <vscale x 8 x double> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x double> %vb, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}
+
+define <vscale x 8 x double> @vfma_vf_nxv8f64_unmasked_commute(<vscale x 8 x double> %va, double %b, <vscale x 8 x double> %vc, i32 zeroext %evl) {
+; CHECK-LABEL: vfma_vf_nxv8f64_unmasked_commute:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, mu
+; CHECK-NEXT:    vfmadd.vf v8, fa0, v16
+; CHECK-NEXT:    ret
+  %elt.head = insertelement <vscale x 8 x double> poison, double %b, i32 0
+  %vb = shufflevector <vscale x 8 x double> %elt.head, <vscale x 8 x double> poison, <vscale x 8 x i32> zeroinitializer
+  %head = insertelement <vscale x 8 x i1> poison, i1 true, i32 0
+  %m = shufflevector <vscale x 8 x i1> %head, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+  %v = call <vscale x 8 x double> @llvm.vp.fma.nxv8f64(<vscale x 8 x double> %vb, <vscale x 8 x double> %va, <vscale x 8 x double> %vc, <vscale x 8 x i1> %m, i32 %evl)
+  ret <vscale x 8 x double> %v
+}