[RISCV] [NFC] add some tests for overloaded intrinsics of FP16

[Chenbing.Zheng]

Summary

These tests for overloaded intrinsics of FP16 is missing in vle/vse and vlse/vsse.

[craig.topper]

LGTM

clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vle.c

 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
 // REQUIRES: riscv-registered-target
-// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v \
-// RUN:   -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
+// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v -target-feature +zfh \
+// RUN:    -target-feature +experimental-zvfh -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
 
 #include <riscv_vector.h>
 
 // CHECK-RV64-LABEL: @test_vle8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
 // CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vle.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]

 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
 // CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vle.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vle64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff, const double *base, size_t vl) {
   return vle64(mask, maskedoff, base, vl);
 }
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16mf4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x half> @llvm.riscv.vle.mask.nxv1f16.i64(<vscale x 1 x half> [[MASKEDOFF:%.*]], <vscale x 1 x half>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 1 x half> [[TMP1]]
+//
+vfloat16mf4_t test_vle16_v_f16mf4_m(vbool64_t mask, vfloat16mf4_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16mf2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x half> @llvm.riscv.vle.mask.nxv2f16.i64(<vscale x 2 x half> [[MASKEDOFF:%.*]], <vscale x 2 x half>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 2 x half> [[TMP1]]
+//
+vfloat16mf2_t test_vle16_v_f16mf2_m(vbool32_t mask, vfloat16mf2_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16m1_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x half> @llvm.riscv.vle.mask.nxv4f16.i64(<vscale x 4 x half> [[MASKEDOFF:%.*]], <vscale x 4 x half>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 4 x half> [[TMP1]]
+//
+vfloat16m1_t test_vle16_v_f16m1_m(vbool16_t mask, vfloat16m1_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16m2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x half> @llvm.riscv.vle.mask.nxv8f16.i64(<vscale x 8 x half> [[MASKEDOFF:%.*]], <vscale x 8 x half>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 8 x half> [[TMP1]]
+//
+vfloat16m2_t test_vle16_v_f16m2_m(vbool8_t mask, vfloat16m2_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16m4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x half> @llvm.riscv.vle.mask.nxv16f16.i64(<vscale x 16 x half> [[MASKEDOFF:%.*]], <vscale x 16 x half>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 16 x half> [[TMP1]]
+//
+vfloat16m4_t test_vle16_v_f16m4_m(vbool4_t mask, vfloat16m4_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vle16_v_f16m8_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x half> @llvm.riscv.vle.mask.nxv32f16.i64(<vscale x 32 x half> [[MASKEDOFF:%.*]], <vscale x 32 x half>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 32 x half> [[TMP1]]
+//
+vfloat16m8_t test_vle16_v_f16m8_m(vbool2_t mask, vfloat16m8_t maskedoff, const _Float16 *base, size_t vl) {
+  return vle16(mask, maskedoff, base, vl);
+}

clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vlse.c

 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
 // REQUIRES: riscv-registered-target
-// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v \
-// RUN:   -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
+// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v -target-feature +zfh \
+// RUN:   -target-feature +experimental-zvfh -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
 
 #include <riscv_vector.h>
 
 // CHECK-RV64-LABEL: @test_vlse8_v_i8mf8_m(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
 // CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vlse.mask.nxv1i8.i64(<vscale x 1 x i8> [[MASKEDOFF:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
 // CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP1]]

 // CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x double> @llvm.riscv.vlse.mask.nxv8f64.i64(<vscale x 8 x double> [[MASKEDOFF:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
 // CHECK-RV64-NEXT:    ret <vscale x 8 x double> [[TMP1]]
 //
 vfloat64m8_t test_vlse64_v_f64m8_m(vbool8_t mask, vfloat64m8_t maskedoff,
                                    const double *base, ptrdiff_t bstride,
                                    size_t vl) {
   return vlse64(mask, maskedoff, base, bstride, vl);
 }
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16mf4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 1 x half> @llvm.riscv.vlse.mask.nxv1f16.i64(<vscale x 1 x half> [[MASKEDOFF:%.*]], <vscale x 1 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 1 x half> [[TMP1]]
+//
+vfloat16mf4_t test_vlse16_v_f16mf4_m(vbool64_t mask, vfloat16mf4_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16mf2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 2 x half> @llvm.riscv.vlse.mask.nxv2f16.i64(<vscale x 2 x half> [[MASKEDOFF:%.*]], <vscale x 2 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 2 x half> [[TMP1]]
+//
+vfloat16mf2_t test_vlse16_v_f16mf2_m(vbool32_t mask, vfloat16mf2_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16m1_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 4 x half> @llvm.riscv.vlse.mask.nxv4f16.i64(<vscale x 4 x half> [[MASKEDOFF:%.*]], <vscale x 4 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 4 x half> [[TMP1]]
+//
+vfloat16m1_t test_vlse16_v_f16m1_m(vbool16_t mask, vfloat16m1_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16m2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 8 x half> @llvm.riscv.vlse.mask.nxv8f16.i64(<vscale x 8 x half> [[MASKEDOFF:%.*]], <vscale x 8 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 8 x half> [[TMP1]]
+//
+vfloat16m2_t test_vlse16_v_f16m2_m(vbool8_t mask, vfloat16m2_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16m4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 16 x half> @llvm.riscv.vlse.mask.nxv16f16.i64(<vscale x 16 x half> [[MASKEDOFF:%.*]], <vscale x 16 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 16 x half> [[TMP1]]
+//
+vfloat16m4_t test_vlse16_v_f16m4_m(vbool4_t mask, vfloat16m4_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vlse16_v_f16m8_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    [[TMP1:%.*]] = call <vscale x 32 x half> @llvm.riscv.vlse.mask.nxv32f16.i64(<vscale x 32 x half> [[MASKEDOFF:%.*]], <vscale x 32 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]], i64 0)
+// CHECK-RV64-NEXT:    ret <vscale x 32 x half> [[TMP1]]
+//
+vfloat16m8_t test_vlse16_v_f16m8_m(vbool2_t mask, vfloat16m8_t maskedoff, const _Float16 *base, ptrdiff_t bstride, size_t vl) {
+  return vlse16(mask, maskedoff, base, bstride, vl);
+}

clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vse.c

 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
 // REQUIRES: riscv-registered-target
-// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v \
-// RUN:   -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
+// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v -target-feature +zfh \
+// RUN:   -target-feature +experimental-zvfh -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
 
 #include <riscv_vector.h>
 
 // CHECK-RV64-LABEL: @test_vse8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
 // CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void

 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i1>*
 // CHECK-RV64-NEXT:    call void @llvm.riscv.vsm.nxv1i1.i64(<vscale x 1 x i1> [[VALUE:%.*]], <vscale x 1 x i1>* [[TMP0]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsm_v_b64(uint8_t *base, vbool64_t value, size_t vl) {
   return vsm(base, value, vl);
 }
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16mf4(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv1f16.i64(<vscale x 1 x half> [[VALUE:%.*]], <vscale x 1 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16mf4(_Float16 *base, vfloat16mf4_t value, size_t vl) {
+  return vse16(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16mf2(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv2f16.i64(<vscale x 2 x half> [[VALUE:%.*]], <vscale x 2 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16mf2(_Float16 *base, vfloat16mf2_t value, size_t vl) {
+  return vse16_v_f16mf2(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m1(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv4f16.i64(<vscale x 4 x half> [[VALUE:%.*]], <vscale x 4 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m1(_Float16 *base, vfloat16m1_t value, size_t vl) {
+  return vse16(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m2(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv8f16.i64(<vscale x 8 x half> [[VALUE:%.*]], <vscale x 8 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m2(_Float16 *base, vfloat16m2_t value, size_t vl) {
+  return vse16(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m4(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv16f16.i64(<vscale x 16 x half> [[VALUE:%.*]], <vscale x 16 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m4(_Float16 *base, vfloat16m4_t value, size_t vl) {
+  return vse16(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m8(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.nxv32f16.i64(<vscale x 32 x half> [[VALUE:%.*]], <vscale x 32 x half>* [[TMP0]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m8(_Float16 *base, vfloat16m8_t value, size_t vl) {
+  return vse16(base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16mf4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv1f16.i64(<vscale x 1 x half> [[VALUE:%.*]], <vscale x 1 x half>* [[TMP0]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16mf4_m(vbool64_t mask, _Float16 *base, vfloat16mf4_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16mf2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv2f16.i64(<vscale x 2 x half> [[VALUE:%.*]], <vscale x 2 x half>* [[TMP0]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16mf2_m(vbool32_t mask, _Float16 *base, vfloat16mf2_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m1_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv4f16.i64(<vscale x 4 x half> [[VALUE:%.*]], <vscale x 4 x half>* [[TMP0]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m1_m(vbool16_t mask, _Float16 *base, vfloat16m1_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv8f16.i64(<vscale x 8 x half> [[VALUE:%.*]], <vscale x 8 x half>* [[TMP0]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m2_m(vbool8_t mask, _Float16 *base, vfloat16m2_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv16f16.i64(<vscale x 16 x half> [[VALUE:%.*]], <vscale x 16 x half>* [[TMP0]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m4_m(vbool4_t mask, _Float16 *base, vfloat16m4_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vse16_v_f16m8_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vse.mask.nxv32f16.i64(<vscale x 32 x half> [[VALUE:%.*]], <vscale x 32 x half>* [[TMP0]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vse16_v_f16m8_m(vbool2_t mask, _Float16 *base, vfloat16m8_t value, size_t vl) {
+  return vse16(mask, base, value, vl);
+}

clang/test/CodeGen/RISCV/rvv-intrinsics-overloaded/vsse.c

 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
 // REQUIRES: riscv-registered-target
-// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v \
-// RUN:   -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
+// RUN: %clang_cc1 -triple riscv64 -target-feature +f -target-feature +d -target-feature +v -target-feature +zfh \
+// RUN:   -target-feature +experimental-zvfh -disable-O0-optnone -emit-llvm %s -o - | opt -S -mem2reg | FileCheck --check-prefix=CHECK-RV64 %s
 
 #include <riscv_vector.h>
 
 // CHECK-RV64-LABEL: @test_vsse8_v_i8mf8(
 // CHECK-RV64-NEXT:  entry:
 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 1 x i8>*
 // CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1i8.i64(<vscale x 1 x i8> [[VALUE:%.*]], <vscale x 1 x i8>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void

 // CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 8 x double>*
 // CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f64.i64(<vscale x 8 x double> [[VALUE:%.*]], <vscale x 8 x double>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
 // CHECK-RV64-NEXT:    ret void
 //
 void test_vsse64_v_f64m8_m(vbool8_t mask, double *base, ptrdiff_t bstride,
                            vfloat64m8_t value, size_t vl) {
   return vsse64(mask, base, bstride, value, vl);
 }
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16mf4(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv1f16.i64(<vscale x 1 x half> [[VALUE:%.*]], <vscale x 1 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16mf4(_Float16 *base, ptrdiff_t bstride, vfloat16mf4_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16mf2(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv2f16.i64(<vscale x 2 x half> [[VALUE:%.*]], <vscale x 2 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16mf2(_Float16 *base, ptrdiff_t bstride, vfloat16mf2_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m1(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv4f16.i64(<vscale x 4 x half> [[VALUE:%.*]], <vscale x 4 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m1(_Float16 *base, ptrdiff_t bstride, vfloat16m1_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m2(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv8f16.i64(<vscale x 8 x half> [[VALUE:%.*]], <vscale x 8 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m2(_Float16 *base, ptrdiff_t bstride, vfloat16m2_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m4(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv16f16.i64(<vscale x 16 x half> [[VALUE:%.*]], <vscale x 16 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m4(_Float16 *base, ptrdiff_t bstride, vfloat16m4_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m8(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.nxv32f16.i64(<vscale x 32 x half> [[VALUE:%.*]], <vscale x 32 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m8(_Float16 *base, ptrdiff_t bstride, vfloat16m8_t value, size_t vl) {
+  return vsse16(base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16mf4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 1 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv1f16.i64(<vscale x 1 x half> [[VALUE:%.*]], <vscale x 1 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 1 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16mf4_m(vbool64_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16mf4_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16mf2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 2 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv2f16.i64(<vscale x 2 x half> [[VALUE:%.*]], <vscale x 2 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 2 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16mf2_m(vbool32_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16mf2_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m1_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 4 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv4f16.i64(<vscale x 4 x half> [[VALUE:%.*]], <vscale x 4 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 4 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m1_m(vbool16_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16m1_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m2_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv8f16.i64(<vscale x 8 x half> [[VALUE:%.*]], <vscale x 8 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 8 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m2_m(vbool8_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16m2_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m4_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 16 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv16f16.i64(<vscale x 16 x half> [[VALUE:%.*]], <vscale x 16 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 16 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m4_m(vbool4_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16m4_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}
+
+// CHECK-RV64-LABEL: @test_vsse16_v_f16m8_m(
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 32 x half>*
+// CHECK-RV64-NEXT:    call void @llvm.riscv.vsse.mask.nxv32f16.i64(<vscale x 32 x half> [[VALUE:%.*]], <vscale x 32 x half>* [[TMP0]], i64 [[BSTRIDE:%.*]], <vscale x 32 x i1> [[MASK:%.*]], i64 [[VL:%.*]])
+// CHECK-RV64-NEXT:    ret void
+//
+void test_vsse16_v_f16m8_m(vbool2_t mask, _Float16 *base, ptrdiff_t bstride, vfloat16m8_t value, size_t vl) {
+  return vsse16(mask, base, bstride, value, vl);
+}