// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// RUN: %clang_cc1 -triple thumbv8.1m.main-none-none-eabi -target-feature +mve.fp -mfloat-abi hard -fallow-half-arguments-and-returns -O0 -disable-O0-optnone -S -emit-llvm -o - %s | opt -S -mem2reg -sroa | FileCheck %s
// RUN: %clang_cc1 -triple thumbv8.1m.main-none-none-eabi -target-feature +mve.fp -mfloat-abi hard -fallow-half-arguments-and-returns -O0 -disable-O0-optnone -DPOLYMORPHIC -S -emit-llvm -o - %s | opt -S -mem2reg -sroa | FileCheck %s

// REQUIRES: aarch64-registered-target || arm-registered-target

#include <arm_mve.h>

// CHECK-LABEL: @test_vmladavaq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 0, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaq_s8(int32_t a, int8x16_t b, int8x16_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_s8(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 0, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaq_s16(int32_t a, int16x8_t b, int16x8_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_s16(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 0, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaq_s32(int32_t a, int32x4_t b, int32x4_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_s32(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 1, i32 0, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavaq_u8(uint32_t a, uint8x16_t b, uint8x16_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_u8(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_u16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 1, i32 0, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavaq_u16(uint32_t a, uint16x8_t b, uint16x8_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_u16(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 1, i32 0, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavaq_u32(uint32_t a, uint32x4_t b, uint32x4_t c) {
#ifdef POLYMORPHIC
  return vmladavaq(a, b, c);
#else
  return vmladavaq_u32(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 0, i32 1, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaxq_s8(int32_t a, int8x16_t b, int8x16_t c) {
#ifdef POLYMORPHIC
  return vmladavaxq(a, b, c);
#else
  return vmladavaxq_s8(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 0, i32 1, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaxq_s16(int32_t a, int16x8_t b, int16x8_t c) {
#ifdef POLYMORPHIC
  return vmladavaxq(a, b, c);
#else
  return vmladavaxq_s16(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 0, i32 1, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavaxq_s32(int32_t a, int32x4_t b, int32x4_t c) {
#ifdef POLYMORPHIC
  return vmladavaxq(a, b, c);
#else
  return vmladavaxq_s32(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 1, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaq_s8(int32_t a, int8x16_t b, int8x16_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaq(a, b, c);
#else
  return vmlsdavaq_s8(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 1, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaq_s16(int32_t a, int16x8_t b, int16x8_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaq(a, b, c);
#else
  return vmlsdavaq_s16(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 1, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaq_s32(int32_t a, int32x4_t b, int32x4_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaq(a, b, c);
#else
  return vmlsdavaq_s32(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 1, i32 1, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaxq_s8(int32_t a, int8x16_t b, int8x16_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaxq(a, b, c);
#else
  return vmlsdavaxq_s8(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 1, i32 1, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaxq_s16(int32_t a, int16x8_t b, int16x8_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaxq(a, b, c);
#else
  return vmlsdavaxq_s16(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 1, i32 1, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavaxq_s32(int32_t a, int32x4_t b, int32x4_t c) {
#ifdef POLYMORPHIC
  return vmlsdavaxq(a, b, c);
#else
  return vmlsdavaxq_s32(a, b, c);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 0, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaq_p_s8(int32_t a, int8x16_t b, int8x16_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_s8(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 0, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaq_p_s16(int32_t a, int16x8_t b, int16x8_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_s16(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 0, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaq_p_s32(int32_t a, int32x4_t b, int32x4_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_s32(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 1, i32 0, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavaq_p_u8(uint32_t a, uint8x16_t b, uint8x16_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_u8(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_u16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 1, i32 0, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavaq_p_u16(uint32_t a, uint16x8_t b, uint16x8_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_u16(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaq_p_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 1, i32 0, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavaq_p_u32(uint32_t a, uint32x4_t b, uint32x4_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaq_p(a, b, c, p);
#else
  return vmladavaq_p_u32(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 0, i32 1, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaxq_p_s8(int32_t a, int8x16_t b, int8x16_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaxq_p(a, b, c, p);
#else
  return vmladavaxq_p_s8(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 0, i32 1, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaxq_p_s16(int32_t a, int16x8_t b, int16x8_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaxq_p(a, b, c, p);
#else
  return vmladavaxq_p_s16(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavaxq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 0, i32 1, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavaxq_p_s32(int32_t a, int32x4_t b, int32x4_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavaxq_p(a, b, c, p);
#else
  return vmladavaxq_p_s32(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 1, i32 0, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaq_p_s8(int32_t a, int8x16_t b, int8x16_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaq_p(a, b, c, p);
#else
  return vmlsdavaq_p_s8(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 1, i32 0, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaq_p_s16(int32_t a, int16x8_t b, int16x8_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaq_p(a, b, c, p);
#else
  return vmlsdavaq_p_s16(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 1, i32 0, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaq_p_s32(int32_t a, int32x4_t b, int32x4_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaq_p(a, b, c, p);
#else
  return vmlsdavaq_p_s32(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 1, i32 1, i32 [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i8> [[C:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaxq_p_s8(int32_t a, int8x16_t b, int8x16_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaxq_p(a, b, c, p);
#else
  return vmlsdavaxq_p_s8(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 1, i32 1, i32 [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i16> [[C:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaxq_p_s16(int32_t a, int16x8_t b, int16x8_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaxq_p(a, b, c, p);
#else
  return vmlsdavaxq_p_s16(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavaxq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 1, i32 1, i32 [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i32> [[C:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavaxq_p_s32(int32_t a, int32x4_t b, int32x4_t c, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavaxq_p(a, b, c, p);
#else
  return vmlsdavaxq_p_s32(a, b, c, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 0, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavq_s8(int8x16_t a, int8x16_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_s8(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 0, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavq_s16(int16x8_t a, int16x8_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_s16(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 0, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavq_s32(int32x4_t a, int32x4_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_s32(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 1, i32 0, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavq_u8(uint8x16_t a, uint8x16_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_u8(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_u16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 1, i32 0, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavq_u16(uint16x8_t a, uint16x8_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_u16(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 1, i32 0, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
uint32_t test_vmladavq_u32(uint32x4_t a, uint32x4_t b) {
#ifdef POLYMORPHIC
  return vmladavq(a, b);
#else
  return vmladavq_u32(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavxq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 0, i32 1, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavxq_s8(int8x16_t a, int8x16_t b) {
#ifdef POLYMORPHIC
  return vmladavxq(a, b);
#else
  return vmladavxq_s8(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavxq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 0, i32 1, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavxq_s16(int16x8_t a, int16x8_t b) {
#ifdef POLYMORPHIC
  return vmladavxq(a, b);
#else
  return vmladavxq_s16(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavxq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 0, i32 1, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmladavxq_s32(int32x4_t a, int32x4_t b) {
#ifdef POLYMORPHIC
  return vmladavxq(a, b);
#else
  return vmladavxq_s32(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 1, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavq_s8(int8x16_t a, int8x16_t b) {
#ifdef POLYMORPHIC
  return vmlsdavq(a, b);
#else
  return vmlsdavq_s8(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 1, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavq_s16(int16x8_t a, int16x8_t b) {
#ifdef POLYMORPHIC
  return vmlsdavq(a, b);
#else
  return vmlsdavq_s16(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 1, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavq_s32(int32x4_t a, int32x4_t b) {
#ifdef POLYMORPHIC
  return vmlsdavq(a, b);
#else
  return vmlsdavq_s32(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v16i8(i32 0, i32 1, i32 1, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavxq_s8(int8x16_t a, int8x16_t b) {
#ifdef POLYMORPHIC
  return vmlsdavxq(a, b);
#else
  return vmlsdavxq_s8(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v8i16(i32 0, i32 1, i32 1, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavxq_s16(int16x8_t a, int16x8_t b) {
#ifdef POLYMORPHIC
  return vmlsdavxq(a, b);
#else
  return vmlsdavxq_s16(a, b);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = call i32 @llvm.arm.mve.vmldava.v4i32(i32 0, i32 1, i32 1, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]])
// CHECK-NEXT:    ret i32 [[TMP0]]
//
int32_t test_vmlsdavxq_s32(int32x4_t a, int32x4_t b) {
#ifdef POLYMORPHIC
  return vmlsdavxq(a, b);
#else
  return vmlsdavxq_s32(a, b);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 0, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavq_p_s8(int8x16_t a, int8x16_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_s8(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 0, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavq_p_s16(int16x8_t a, int16x8_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_s16(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 0, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavq_p_s32(int32x4_t a, int32x4_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_s32(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 1, i32 0, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavq_p_u8(uint8x16_t a, uint8x16_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_u8(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_u16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 1, i32 0, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavq_p_u16(uint16x8_t a, uint16x8_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_u16(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavq_p_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 1, i32 0, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
uint32_t test_vmladavq_p_u32(uint32x4_t a, uint32x4_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavq_p(a, b, p);
#else
  return vmladavq_p_u32(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavxq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 0, i32 1, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavxq_p_s8(int8x16_t a, int8x16_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavxq_p(a, b, p);
#else
  return vmladavxq_p_s8(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavxq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 0, i32 1, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavxq_p_s16(int16x8_t a, int16x8_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavxq_p(a, b, p);
#else
  return vmladavxq_p_s16(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmladavxq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 0, i32 1, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmladavxq_p_s32(int32x4_t a, int32x4_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmladavxq_p(a, b, p);
#else
  return vmladavxq_p_s32(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 1, i32 0, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavq_p_s8(int8x16_t a, int8x16_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavq_p(a, b, p);
#else
  return vmlsdavq_p_s8(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 1, i32 0, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavq_p_s16(int16x8_t a, int16x8_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavq_p(a, b, p);
#else
  return vmlsdavq_p_s16(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 1, i32 0, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavq_p_s32(int32x4_t a, int32x4_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavq_p(a, b, p);
#else
  return vmlsdavq_p_s32(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_p_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v16i8.v16i1(i32 0, i32 1, i32 1, i32 0, <16 x i8> [[A:%.*]], <16 x i8> [[B:%.*]], <16 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavxq_p_s8(int8x16_t a, int8x16_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavxq_p(a, b, p);
#else
  return vmlsdavxq_p_s8(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_p_s16(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v8i16.v8i1(i32 0, i32 1, i32 1, i32 0, <8 x i16> [[A:%.*]], <8 x i16> [[B:%.*]], <8 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavxq_p_s16(int16x8_t a, int16x8_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavxq_p(a, b, p);
#else
  return vmlsdavxq_p_s16(a, b, p);
#endif
}

// CHECK-LABEL: @test_vmlsdavxq_p_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32
// CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]])
// CHECK-NEXT:    [[TMP2:%.*]] = call i32 @llvm.arm.mve.vmldava.predicated.v4i32.v4i1(i32 0, i32 1, i32 1, i32 0, <4 x i32> [[A:%.*]], <4 x i32> [[B:%.*]], <4 x i1> [[TMP1]])
// CHECK-NEXT:    ret i32 [[TMP2]]
//
int32_t test_vmlsdavxq_p_s32(int32x4_t a, int32x4_t b, mve_pred16_t p) {
#ifdef POLYMORPHIC
  return vmlsdavxq_p(a, b, p);
#else
  return vmlsdavxq_p_s32(a, b, p);
#endif
}