diff --git a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
index 287800b1fb224..edf4ae06977d6 100644
--- a/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
+++ b/llvm/lib/Target/RISCV/RISCVISelLowering.cpp
@@ -1318,16 +1318,6 @@ bool RISCVTargetLowering::shouldSinkOperands(
         case Intrinsic::fma:
         case Intrinsic::vp_fma:
           return Operand == 0 || Operand == 1;
-        // FIXME: Our patterns can only match vx/vf instructions when the splat
-        // it on the RHS, because TableGen doesn't recognize our VP operations
-        // as commutative.
-        case Intrinsic::vp_add:
-        case Intrinsic::vp_mul:
-        case Intrinsic::vp_and:
-        case Intrinsic::vp_or:
-        case Intrinsic::vp_xor:
-        case Intrinsic::vp_fadd:
-        case Intrinsic::vp_fmul:
         case Intrinsic::vp_shl:
         case Intrinsic::vp_lshr:
         case Intrinsic::vp_ashr:
@@ -1336,8 +1326,15 @@ bool RISCVTargetLowering::shouldSinkOperands(
         case Intrinsic::vp_urem:
         case Intrinsic::vp_srem:
           return Operand == 1;
-        // ... with the exception of vp.sub/vp.fsub/vp.fdiv, which have
-        // explicit patterns for both LHS and RHS (as 'vr' versions).
+        // These intrinsics are commutative.
+        case Intrinsic::vp_add:
+        case Intrinsic::vp_mul:
+        case Intrinsic::vp_and:
+        case Intrinsic::vp_or:
+        case Intrinsic::vp_xor:
+        case Intrinsic::vp_fadd:
+        case Intrinsic::vp_fmul:
+        // These intrinsics have 'vr' versions.
         case Intrinsic::vp_sub:
         case Intrinsic::vp_fsub:
         case Intrinsic::vp_fdiv:
diff --git a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
index 8542430f51f3c..9319bd4a5a081 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
@@ -3025,25 +3025,21 @@ for.cond.cleanup:                                 ; preds = %vector.body
   ret void
 }
 
-; FIXME: This doesn't match against vadd.vx because our patterns aren't
-; commutative.
-
 define void @sink_splat_vp_add_commute(i32* nocapture %a, i32 signext %x, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_add_commute:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    li a3, 1024
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
-; CHECK-NEXT:    vmv.v.x v8, a1
-; CHECK-NEXT:    li a1, 1024
 ; CHECK-NEXT:  .LBB48_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    vle32.v v9, (a0)
+; CHECK-NEXT:    vle32.v v8, (a0)
 ; CHECK-NEXT:    vsetvli zero, a2, e32, m1, ta, ma
-; CHECK-NEXT:    vadd.vv v9, v8, v9, v0.t
+; CHECK-NEXT:    vadd.vx v8, v8, a1, v0.t
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
-; CHECK-NEXT:    vse32.v v9, (a0)
-; CHECK-NEXT:    addi a1, a1, -4
+; CHECK-NEXT:    vse32.v v8, (a0)
+; CHECK-NEXT:    addi a3, a3, -4
 ; CHECK-NEXT:    addi a0, a0, 16
-; CHECK-NEXT:    bnez a1, .LBB48_1
+; CHECK-NEXT:    bnez a3, .LBB48_1
 ; CHECK-NEXT:  # %bb.2: # %for.cond.cleanup
 ; CHECK-NEXT:    ret
 entry: