diff --git a/clang/include/clang/Basic/TargetBuiltins.h b/clang/include/clang/Basic/TargetBuiltins.h
--- a/clang/include/clang/Basic/TargetBuiltins.h
+++ b/clang/include/clang/Basic/TargetBuiltins.h
@@ -289,6 +289,7 @@
     bool isGatherPrefetch() const { return Flags & IsGatherPrefetch; }
     bool isReverseUSDOT() const { return Flags & ReverseUSDOT; }
     bool isReverseMergeAnyBinOp() const { return Flags & ReverseMergeAnyBinOp; }
+    bool isReverseMergeAnyAccOp() const { return Flags & ReverseMergeAnyAccOp; }
     bool isUndef() const { return Flags & IsUndef; }
     bool isTupleCreate() const { return Flags & IsTupleCreate; }
     bool isTupleGet() const { return Flags & IsTupleGet; }
diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -206,6 +206,7 @@
 def IsTupleGet                : FlagType<0x200000000>;
 def IsTupleSet                : FlagType<0x400000000>;
 def ReverseMergeAnyBinOp      : FlagType<0x800000000>; // e.g. Implement SUBR_X using SUB_X.
+def ReverseMergeAnyAccOp      : FlagType<0x1000000000>; // e.g. Implement MSB_X using MLS_X.
 
 // These must be kept in sync with the flags in include/clang/Basic/TargetBuiltins.h
 class ImmCheckType<int val> {
@@ -788,20 +789,20 @@
 
 //------------------------------------------------------------------------------
 
-multiclass SInstZPZZZ<string name, string types, string intrinsic, list<FlagType> flags=[]> {
-  def _M   : SInst<name # "[_{d}]",   "dPddd", types, MergeOp1,  intrinsic, flags>;
-  def _X   : SInst<name # "[_{d}]",   "dPddd", types, MergeAny,  intrinsic, flags>;
-  def _Z   : SInst<name # "[_{d}]",   "dPddd", types, MergeZero, intrinsic, flags>;
+multiclass SInstZPZZZ<string name, string types, string m_intrinsic, string x_intrinsic, list<FlagType> flags=[]> {
+  def _M   : SInst<name # "[_{d}]",   "dPddd", types, MergeOp1,  m_intrinsic, flags>;
+  def _X   : SInst<name # "[_{d}]",   "dPddd", types, MergeAny,  x_intrinsic, flags>;
+  def _Z   : SInst<name # "[_{d}]",   "dPddd", types, MergeZero, m_intrinsic, flags>;
 
-  def _N_M : SInst<name # "[_n_{d}]", "dPdda", types, MergeOp1,  intrinsic, flags>;
-  def _N_X : SInst<name # "[_n_{d}]", "dPdda", types, MergeAny,  intrinsic, flags>;
-  def _N_Z : SInst<name # "[_n_{d}]", "dPdda", types, MergeZero, intrinsic, flags>;
+  def _N_M : SInst<name # "[_n_{d}]", "dPdda", types, MergeOp1,  m_intrinsic, flags>;
+  def _N_X : SInst<name # "[_n_{d}]", "dPdda", types, MergeAny,  x_intrinsic, flags>;
+  def _N_Z : SInst<name # "[_n_{d}]", "dPdda", types, MergeZero, m_intrinsic, flags>;
 }
 
-defm SVMAD : SInstZPZZZ<"svmad", "csilUcUsUiUl", "aarch64_sve_mad">;
-defm SVMLA : SInstZPZZZ<"svmla", "csilUcUsUiUl", "aarch64_sve_mla">;
-defm SVMLS : SInstZPZZZ<"svmls", "csilUcUsUiUl", "aarch64_sve_mls">;
-defm SVMSB : SInstZPZZZ<"svmsb", "csilUcUsUiUl", "aarch64_sve_msb">;
+defm SVMAD : SInstZPZZZ<"svmad", "csilUcUsUiUl", "aarch64_sve_mad", "aarch64_sve_mad">;
+defm SVMLA : SInstZPZZZ<"svmla", "csilUcUsUiUl", "aarch64_sve_mla", "aarch64_sve_mla">;
+defm SVMLS : SInstZPZZZ<"svmls", "csilUcUsUiUl", "aarch64_sve_mls", "aarch64_sve_mls">;
+defm SVMSB : SInstZPZZZ<"svmsb", "csilUcUsUiUl", "aarch64_sve_msb", "aarch64_sve_msb">;
 
 //------------------------------------------------------------------------------
 
@@ -980,18 +981,18 @@
 defm SVABS_F : SInstZPZ<"svabs", "hfd", "aarch64_sve_fabs">;
 defm SVNEG_F : SInstZPZ<"svneg", "hfd", "aarch64_sve_fneg">;
 
-defm SVABD_F  : SInstZPZZ<"svabd",  "hfd", "aarch64_sve_fabd",   "aarch64_sve_fabd">;
-defm SVADD_F  : SInstZPZZ<"svadd",  "hfd", "aarch64_sve_fadd",   "aarch64_sve_fadd">;
-defm SVDIV_F  : SInstZPZZ<"svdiv",  "hfd", "aarch64_sve_fdiv",   "aarch64_sve_fdiv">;
-defm SVDIVR_F : SInstZPZZ<"svdivr", "hfd", "aarch64_sve_fdivr",  "aarch64_sve_fdivr">;
-defm SVMAX_F  : SInstZPZZ<"svmax",  "hfd", "aarch64_sve_fmax",   "aarch64_sve_fmax">;
-defm SVMAXNM  : SInstZPZZ<"svmaxnm","hfd", "aarch64_sve_fmaxnm", "aarch64_sve_fmaxnm">;
-defm SVMIN_F  : SInstZPZZ<"svmin",  "hfd", "aarch64_sve_fmin",   "aarch64_sve_fmin">;
-defm SVMINNM  : SInstZPZZ<"svminnm","hfd", "aarch64_sve_fminnm", "aarch64_sve_fminnm">;
-defm SVMUL_F  : SInstZPZZ<"svmul",  "hfd", "aarch64_sve_fmul",   "aarch64_sve_fmul">;
-defm SVMULX   : SInstZPZZ<"svmulx", "hfd", "aarch64_sve_fmulx",  "aarch64_sve_fmulx">;
-defm SVSUB_F  : SInstZPZZ<"svsub",  "hfd", "aarch64_sve_fsub",   "aarch64_sve_fsub">;
-defm SVSUBR_F : SInstZPZZ<"svsubr", "hfd", "aarch64_sve_fsubr",  "aarch64_sve_fsubr">;
+defm SVABD_F  : SInstZPZZ<"svabd",  "hfd", "aarch64_sve_fabd",   "aarch64_sve_fabd_u">;
+defm SVADD_F  : SInstZPZZ<"svadd",  "hfd", "aarch64_sve_fadd",   "aarch64_sve_fadd_u">;
+defm SVDIV_F  : SInstZPZZ<"svdiv",  "hfd", "aarch64_sve_fdiv",   "aarch64_sve_fdiv_u">;
+defm SVDIVR_F : SInstZPZZ<"svdivr", "hfd", "aarch64_sve_fdivr",  "aarch64_sve_fdiv_u", [ReverseMergeAnyBinOp]>;
+defm SVMAX_F  : SInstZPZZ<"svmax",  "hfd", "aarch64_sve_fmax",   "aarch64_sve_fmax_u">;
+defm SVMAXNM  : SInstZPZZ<"svmaxnm","hfd", "aarch64_sve_fmaxnm", "aarch64_sve_fmaxnm_u">;
+defm SVMIN_F  : SInstZPZZ<"svmin",  "hfd", "aarch64_sve_fmin",   "aarch64_sve_fmin_u">;
+defm SVMINNM  : SInstZPZZ<"svminnm","hfd", "aarch64_sve_fminnm", "aarch64_sve_fminnm_u">;
+defm SVMUL_F  : SInstZPZZ<"svmul",  "hfd", "aarch64_sve_fmul",   "aarch64_sve_fmul_u">;
+defm SVMULX   : SInstZPZZ<"svmulx", "hfd", "aarch64_sve_fmulx",  "aarch64_sve_fmulx_u">;
+defm SVSUB_F  : SInstZPZZ<"svsub",  "hfd", "aarch64_sve_fsub",   "aarch64_sve_fsub_u">;
+defm SVSUBR_F : SInstZPZZ<"svsubr", "hfd", "aarch64_sve_fsubr",  "aarch64_sve_fsub_u", [ReverseMergeAnyBinOp]>;
 
 defm SVRECPX : SInstZPZ<"svrecpx", "hfd", "aarch64_sve_frecpx">;
 defm SVRINTA : SInstZPZ<"svrinta", "hfd", "aarch64_sve_frinta">;
@@ -1016,14 +1017,14 @@
 def SVSCALE_N_X : SInst<"svscale[_n_{d}]", "dPdK", "hfd", MergeAny,  "aarch64_sve_fscale">;
 def SVSCALE_N_Z : SInst<"svscale[_n_{d}]", "dPdK", "hfd", MergeZero, "aarch64_sve_fscale">;
 
-defm SVMAD_F  : SInstZPZZZ<"svmad",  "hfd", "aarch64_sve_fmad">;
-defm SVMLA_F  : SInstZPZZZ<"svmla",  "hfd", "aarch64_sve_fmla">;
-defm SVMLS_F  : SInstZPZZZ<"svmls",  "hfd", "aarch64_sve_fmls">;
-defm SVMSB_F  : SInstZPZZZ<"svmsb",  "hfd", "aarch64_sve_fmsb">;
-defm SVNMAD_F : SInstZPZZZ<"svnmad", "hfd", "aarch64_sve_fnmad">;
-defm SVNMLA_F : SInstZPZZZ<"svnmla", "hfd", "aarch64_sve_fnmla">;
-defm SVNMLS_F : SInstZPZZZ<"svnmls", "hfd", "aarch64_sve_fnmls">;
-defm SVNMSB_F : SInstZPZZZ<"svnmsb", "hfd", "aarch64_sve_fnmsb">;
+defm SVMAD_F  : SInstZPZZZ<"svmad",  "hfd", "aarch64_sve_fmad",  "aarch64_sve_fmla_u",  [ReverseMergeAnyAccOp]>;
+defm SVMLA_F  : SInstZPZZZ<"svmla",  "hfd", "aarch64_sve_fmla",  "aarch64_sve_fmla_u">;
+defm SVMLS_F  : SInstZPZZZ<"svmls",  "hfd", "aarch64_sve_fmls",  "aarch64_sve_fmls_u">;
+defm SVMSB_F  : SInstZPZZZ<"svmsb",  "hfd", "aarch64_sve_fmsb",  "aarch64_sve_fmls_u",  [ReverseMergeAnyAccOp]>;
+defm SVNMAD_F : SInstZPZZZ<"svnmad", "hfd", "aarch64_sve_fnmad", "aarch64_sve_fnmla_u", [ReverseMergeAnyAccOp]>;
+defm SVNMLA_F : SInstZPZZZ<"svnmla", "hfd", "aarch64_sve_fnmla", "aarch64_sve_fnmla_u">;
+defm SVNMLS_F : SInstZPZZZ<"svnmls", "hfd", "aarch64_sve_fnmls", "aarch64_sve_fnmls_u">;
+defm SVNMSB_F : SInstZPZZZ<"svnmsb", "hfd", "aarch64_sve_fnmsb", "aarch64_sve_fnmls_u", [ReverseMergeAnyAccOp]>;
 
 def SVCADD_M : SInst<"svcadd[_{d}]", "dPddi",  "hfd", MergeOp1,  "aarch64_sve_fcadd", [], [ImmCheck<3, ImmCheckComplexRot90_270>]>;
 def SVCADD_X : SInst<"svcadd[_{d}]", "dPddi",  "hfd", MergeAny,  "aarch64_sve_fcadd", [], [ImmCheck<3, ImmCheckComplexRot90_270>]>;
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -9522,6 +9522,9 @@
     else if (TypeFlags.isReverseMergeAnyBinOp() &&
              TypeFlags.getMergeType() == SVETypeFlags::MergeAny)
       std::swap(Ops[1], Ops[2]);
+    else if (TypeFlags.isReverseMergeAnyAccOp() &&
+             TypeFlags.getMergeType() == SVETypeFlags::MergeAny)
+      std::swap(Ops[1], Ops[3]);
 
     // Predicated intrinsics with _z suffix need a select w/ zeroinitializer.
     if (TypeFlags.getMergeType() == SVETypeFlags::MergeZero) {
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_abd.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_abd.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_abd.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_abd.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svabd_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svabd_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svabd_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svabd_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svabd_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svabd_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svabd_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svabd_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svabd_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svabd_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fabd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svabd_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svabd_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fabd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svabd_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svabd_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fabd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svabd_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_add.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_add.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_add.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_add.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svadd_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svadd_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svadd_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svadd_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svadd_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svadd_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svadd_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svadd_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svadd_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svadd_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fadd.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svadd_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svadd_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fadd.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svadd_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svadd_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fadd.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svadd_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_div.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_div.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_div.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_div.c
@@ -597,13 +597,13 @@
 // CHECK-LABEL: @test_svdiv_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svdiv_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svdiv_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -614,13 +614,13 @@
 // CHECK-LABEL: @test_svdiv_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svdiv_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svdiv_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -631,13 +631,13 @@
 // CHECK-LABEL: @test_svdiv_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svdiv_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svdiv_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -782,7 +782,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svdiv_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -790,7 +790,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svdiv_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -803,7 +803,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svdiv_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -811,7 +811,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svdiv_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -824,7 +824,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svdiv_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -832,7 +832,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svdiv_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_divr.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_divr.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_divr.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_divr.c
@@ -597,13 +597,13 @@
 // CHECK-LABEL: @test_svdivr_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdivr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svdivr_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdivr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svdivr_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -614,13 +614,13 @@
 // CHECK-LABEL: @test_svdivr_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdivr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svdivr_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdivr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svdivr_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -631,13 +631,13 @@
 // CHECK-LABEL: @test_svdivr_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdivr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svdivr_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdivr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svdivr_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -782,7 +782,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdivr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svdivr_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -790,7 +790,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdivr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fdiv.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svdivr_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -803,7 +803,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdivr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svdivr_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -811,7 +811,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdivr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fdiv.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svdivr_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -824,7 +824,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdivr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svdivr_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -832,7 +832,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdivr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fdiv.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svdivr_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mad.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mad.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mad.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mad.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmad_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmad_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmad_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmad_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmad_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmad_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmad_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmad_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmad_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmad_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmad_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmad_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmad_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmad_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmad_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_max.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_max.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_max.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_max.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmax_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmax_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmax_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmax_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmax_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmax_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmax_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmax_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmax_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmax_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmax.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmax_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmax_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmax.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmax_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmax_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmax.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmax_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_maxnm.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_maxnm.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_maxnm.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_maxnm.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svmaxnm_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmaxnm_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmaxnm_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svmaxnm_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmaxnm_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmaxnm_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svmaxnm_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmaxnm_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmaxnm_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svmaxnm_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmaxnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmaxnm_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svmaxnm_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmaxnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmaxnm_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svmaxnm_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmaxnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmaxnm_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_min.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_min.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_min.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_min.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmin_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmin_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmin_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmin_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmin_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmin_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmin_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmin_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmin_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmin_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmin.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmin_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmin_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmin.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmin_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmin_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmin.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmin_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_minnm.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_minnm.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_minnm.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_minnm.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svminnm_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svminnm_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svminnm_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svminnm_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svminnm_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svminnm_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svminnm_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svminnm_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svminnm_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svminnm_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fminnm.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svminnm_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svminnm_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fminnm.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svminnm_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z20test_svminnm_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fminnm.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svminnm_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mla.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mla.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mla.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mla.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmla_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmla_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmla_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmla_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmla_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmla_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmla_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmla_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmla_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmla_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmla_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmla_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmla_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmla_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmla_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mls.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mls.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mls.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mls.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmls_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmls_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmls_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmls_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmls_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmls_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmls_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmls_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmls_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmls_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmls_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmls_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmls_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmls_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmls_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_msb.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_msb.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_msb.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_msb.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmsb_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmsb_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmsb_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmsb_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmsb_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmsb_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmsb_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmsb_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmsb_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmsb_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmsb_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmsb_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmsb_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmsb_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmsb_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mul.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mul.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mul.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mul.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svmul_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmul_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmul_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svmul_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmul_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmul_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svmul_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svmul_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmul_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmul_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmul.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmul_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmul_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmul.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmul_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svmul_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmul.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmul_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mulx.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mulx.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mulx.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_mulx.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svmulx_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svmulx_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmulx_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svmulx_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svmulx_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmulx_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svmulx_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svmulx_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmulx_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svmulx_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fmulx.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svmulx_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svmulx_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fmulx.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svmulx_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svmulx_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fmulx.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svmulx_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmad.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmad.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmad.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmad.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svnmad_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmad_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmad_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svnmad_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmad_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmad_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svnmad_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmad_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmad_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmad_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmad.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmad_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmad_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmad.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmad_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmad_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmad.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmad_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmla.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmla.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmla.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmla.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svnmla_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmla_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmla_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svnmla_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmla_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmla_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svnmla_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmla_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmla_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmla_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmla.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmla_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmla_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmla.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmla_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmla_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmla.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmla_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmls.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmls.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmls.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmls.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svnmls_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmls_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmls_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svnmls_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmls_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmls_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svnmls_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmls_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmls_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmls_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmls_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmls_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmls_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmls_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmls_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmsb.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmsb.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmsb.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_nmsb.c
@@ -125,13 +125,13 @@
 // CHECK-LABEL: @test_svnmsb_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmsb_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP3:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmsb_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, svfloat16_t op3)
@@ -142,13 +142,13 @@
 // CHECK-LABEL: @test_svnmsb_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmsb_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP3:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmsb_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, svfloat32_t op3)
@@ -159,13 +159,13 @@
 // CHECK-LABEL: @test_svnmsb_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svnmsb_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP3:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP3:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmsb_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, svfloat64_t op3)
@@ -310,7 +310,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmsb_n_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_tDh(
@@ -318,7 +318,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmsb.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fnmls.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svnmsb_n_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2, float16_t op3)
@@ -331,7 +331,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmsb_n_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_tf(
@@ -339,7 +339,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmsb.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fnmls.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svnmsb_n_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2, float32_t op3)
@@ -352,7 +352,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svnmsb_n_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_td(
@@ -360,7 +360,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP3:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmsb.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fnmls.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svnmsb_n_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2, float64_t op3)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_sub.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_sub.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_sub.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_sub.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svsub_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svsub_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svsub_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svsub_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svsub_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svsub_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svsub_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z16test_svsub_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svsub_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svsub_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svsub_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svsub_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svsub_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z18test_svsub_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svsub_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)
diff --git a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_subr.c b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_subr.c
--- a/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_subr.c
+++ b/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_subr.c
@@ -1045,13 +1045,13 @@
 // CHECK-LABEL: @test_svsubr_f16_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsubr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svsubr_f16_xu10__SVBool_tu13__SVFloat16_tu13__SVFloat16_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsubr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP2:%.*]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svsubr_f16_x(svbool_t pg, svfloat16_t op1, svfloat16_t op2)
@@ -1062,13 +1062,13 @@
 // CHECK-LABEL: @test_svsubr_f32_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsubr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svsubr_f32_xu10__SVBool_tu13__SVFloat32_tu13__SVFloat32_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsubr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP2:%.*]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svsubr_f32_x(svbool_t pg, svfloat32_t op1, svfloat32_t op2)
@@ -1079,13 +1079,13 @@
 // CHECK-LABEL: @test_svsubr_f64_x(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsubr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z17test_svsubr_f64_xu10__SVBool_tu13__SVFloat64_tu13__SVFloat64_t(
 // CPP-CHECK-NEXT:  entry:
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsubr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[OP2:%.*]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP2:%.*]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svsubr_f64_x(svbool_t pg, svfloat64_t op1, svfloat64_t op2)
@@ -1230,7 +1230,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsubr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svsubr_n_f16_xu10__SVBool_tu13__SVFloat16_tDh(
@@ -1238,7 +1238,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 8 x half> poison, half [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 8 x half> [[DOTSPLATINSERT]], <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsubr.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[OP1:%.*]], <vscale x 8 x half> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.aarch64.sve.fsub.u.nxv8f16(<vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> [[DOTSPLAT]], <vscale x 8 x half> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP1]]
 //
 svfloat16_t test_svsubr_n_f16_x(svbool_t pg, svfloat16_t op1, float16_t op2)
@@ -1251,7 +1251,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsubr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svsubr_n_f32_xu10__SVBool_tu13__SVFloat32_tf(
@@ -1259,7 +1259,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x float> poison, float [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x float> [[DOTSPLATINSERT]], <vscale x 4 x float> poison, <vscale x 4 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsubr.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[OP1:%.*]], <vscale x 4 x float> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.fsub.u.nxv4f32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> [[DOTSPLAT]], <vscale x 4 x float> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
 //
 svfloat32_t test_svsubr_n_f32_x(svbool_t pg, svfloat32_t op1, float32_t op2)
@@ -1272,7 +1272,7 @@
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsubr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP1:%.*]])
 // CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 // CPP-CHECK-LABEL: @_Z19test_svsubr_n_f64_xu10__SVBool_tu13__SVFloat64_td(
@@ -1280,7 +1280,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
 // CPP-CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x double> poison, double [[OP2:%.*]], i64 0
 // CPP-CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x double> [[DOTSPLATINSERT]], <vscale x 2 x double> poison, <vscale x 2 x i32> zeroinitializer
-// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsubr.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[OP1:%.*]], <vscale x 2 x double> [[DOTSPLAT]])
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.fsub.u.nxv2f64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> [[DOTSPLAT]], <vscale x 2 x double> [[OP1:%.*]])
 // CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
 //
 svfloat64_t test_svsubr_n_f64_x(svbool_t pg, svfloat64_t op1, float64_t op2)