// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -no-opaque-pointers -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -no-opaque-pointers -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -no-opaque-pointers -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
// RUN: %clang_cc1 -no-opaque-pointers -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
// RUN: %clang_cc1 -no-opaque-pointers -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns -S -O1 -Werror -o /dev/null %s
#include <arm_sve.h>

#ifdef SVE_OVERLOADED_FORMS
// A simple used,unused... macro, long enough to represent any SVE builtin.
#define SVE_ACLE_FUNC(A1,A2_UNUSED,A3,A4_UNUSED) A1##A3
#else
#define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
#endif

// CHECK-LABEL: @test_svld1_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP0]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z13test_svld1_s8u10__SVBool_tPKa(
// CPP-CHECK-NEXT:  entry:
// CPP-CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP0]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
//
svint8_t test_svld1_s8(svbool_t pg, const int8_t *base)
{
  return SVE_ACLE_FUNC(svld1,_s8,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_s16(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_s16u10__SVBool_tPKs(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP2]]
//
svint16_t test_svld1_s16(svbool_t pg, const int16_t *base)
{
  return SVE_ACLE_FUNC(svld1,_s16,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_s32(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_s32u10__SVBool_tPKi(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
svint32_t test_svld1_s32(svbool_t pg, const int32_t *base)
{
  return SVE_ACLE_FUNC(svld1,_s32,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_s64(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_s64u10__SVBool_tPKl(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
svint64_t test_svld1_s64(svbool_t pg, const int64_t *base)
{
  return SVE_ACLE_FUNC(svld1,_s64,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP0]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z13test_svld1_u8u10__SVBool_tPKh(
// CPP-CHECK-NEXT:  entry:
// CPP-CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP0]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CPP-CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP1]]
//
svuint8_t test_svld1_u8(svbool_t pg, const uint8_t *base)
{
  return SVE_ACLE_FUNC(svld1,_u8,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_u16(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_u16u10__SVBool_tPKt(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP2]]
//
svuint16_t test_svld1_u16(svbool_t pg, const uint16_t *base)
{
  return SVE_ACLE_FUNC(svld1,_u16,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_u32(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_u32u10__SVBool_tPKj(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
svuint32_t test_svld1_u32(svbool_t pg, const uint32_t *base)
{
  return SVE_ACLE_FUNC(svld1,_u32,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_u64(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_u64u10__SVBool_tPKm(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
svuint64_t test_svld1_u64(svbool_t pg, const uint64_t *base)
{
  return SVE_ACLE_FUNC(svld1,_u64,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_f16(
// 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:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0nxv8f16(<vscale x 8 x half>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer), !tbaa [[TBAA11:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 8 x half> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_f16u10__SVBool_tPKDh(
// 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:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0nxv8f16(<vscale x 8 x half>* [[TMP1]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer), !tbaa [[TBAA11:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP2]]
//
svfloat16_t test_svld1_f16(svbool_t pg, const float16_t *base)
{
  return SVE_ACLE_FUNC(svld1,_f16,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_f32(
// 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:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0nxv4f32(<vscale x 4 x float>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer), !tbaa [[TBAA13:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_f32u10__SVBool_tPKf(
// 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:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0nxv4f32(<vscale x 4 x float>* [[TMP1]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer), !tbaa [[TBAA13:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP2]]
//
svfloat32_t test_svld1_f32(svbool_t pg, const float32_t *base)
{
  return SVE_ACLE_FUNC(svld1,_f32,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_f64(
// 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:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0nxv2f64(<vscale x 2 x double>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer), !tbaa [[TBAA15:![0-9]+]]
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z14test_svld1_f64u10__SVBool_tPKd(
// 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:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0nxv2f64(<vscale x 2 x double>* [[TMP1]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer), !tbaa [[TBAA15:![0-9]+]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP2]]
//
svfloat64_t test_svld1_f64(svbool_t pg, const float64_t *base)
{
  return SVE_ACLE_FUNC(svld1,_f64,,)(pg, base);
}

// CHECK-LABEL: @test_svld1_vnum_s8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP1:%.*]] = getelementptr <vscale x 16 x i8>, <vscale x 16 x i8>* [[TMP0]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP3]]
//
// CPP-CHECK-LABEL: @_Z18test_svld1_vnum_s8u10__SVBool_tPKal(
// CPP-CHECK-NEXT:  entry:
// CPP-CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP1:%.*]] = getelementptr <vscale x 16 x i8>, <vscale x 16 x i8>* [[TMP0]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CPP-CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP3]]
//
svint8_t test_svld1_vnum_s8(svbool_t pg, const int8_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_s8,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_s16(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x i16>, <vscale x 8 x i16>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16* [[TMP2]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s16u10__SVBool_tPKsl(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x i16>, <vscale x 8 x i16>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16* [[TMP2]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP4]]
//
svint16_t test_svld1_vnum_s16(svbool_t pg, const int16_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_s16,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_s32(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x i32>, <vscale x 4 x i32>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i32* [[TMP2]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s32u10__SVBool_tPKil(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x i32>, <vscale x 4 x i32>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i32* [[TMP2]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP4]]
//
svint32_t test_svld1_vnum_s32(svbool_t pg, const int32_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_s32,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_s64(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x i64>, <vscale x 2 x i64>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i64* [[TMP2]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s64u10__SVBool_tPKll(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x i64>, <vscale x 2 x i64>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i64* [[TMP2]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP4]]
//
svint64_t test_svld1_vnum_s64(svbool_t pg, const int64_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_s64,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_u8(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP1:%.*]] = getelementptr <vscale x 16 x i8>, <vscale x 16 x i8>* [[TMP0]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <vscale x 16 x i8>*
// CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP3]]
//
// CPP-CHECK-LABEL: @_Z18test_svld1_vnum_u8u10__SVBool_tPKhl(
// CPP-CHECK-NEXT:  entry:
// CPP-CHECK-NEXT:    [[TMP0:%.*]] = bitcast i8* [[BASE:%.*]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP1:%.*]] = getelementptr <vscale x 16 x i8>, <vscale x 16 x i8>* [[TMP0]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[TMP1]] to <vscale x 16 x i8>*
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0nxv16i8(<vscale x 16 x i8>* [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer), !tbaa [[TBAA2]]
// CPP-CHECK-NEXT:    ret <vscale x 16 x i8> [[TMP3]]
//
svuint8_t test_svld1_vnum_u8(svbool_t pg, const uint8_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_u8,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_u16(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x i16>, <vscale x 8 x i16>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16* [[TMP2]] to <vscale x 8 x i16>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u16u10__SVBool_tPKtl(
// 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:%.*]] = bitcast i16* [[BASE:%.*]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x i16>, <vscale x 8 x i16>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i16* [[TMP2]] to <vscale x 8 x i16>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0nxv8i16(<vscale x 8 x i16>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer), !tbaa [[TBAA5]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x i16> [[TMP4]]
//
svuint16_t test_svld1_vnum_u16(svbool_t pg, const uint16_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_u16,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_u32(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x i32>, <vscale x 4 x i32>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i32* [[TMP2]] to <vscale x 4 x i32>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u32u10__SVBool_tPKjl(
// 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:%.*]] = bitcast i32* [[BASE:%.*]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x i32>, <vscale x 4 x i32>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i32* [[TMP2]] to <vscale x 4 x i32>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0nxv4i32(<vscale x 4 x i32>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer), !tbaa [[TBAA7]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP4]]
//
svuint32_t test_svld1_vnum_u32(svbool_t pg, const uint32_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_u32,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_u64(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x i64>, <vscale x 2 x i64>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast i64* [[TMP2]] to <vscale x 2 x i64>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u64u10__SVBool_tPKml(
// 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:%.*]] = bitcast i64* [[BASE:%.*]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x i64>, <vscale x 2 x i64>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast i64* [[TMP2]] to <vscale x 2 x i64>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0nxv2i64(<vscale x 2 x i64>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer), !tbaa [[TBAA9]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP4]]
//
svuint64_t test_svld1_vnum_u64(svbool_t pg, const uint64_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_u64,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_f16(
// 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:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x half>, <vscale x 8 x half>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast half* [[TMP2]] to <vscale x 8 x half>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0nxv8f16(<vscale x 8 x half>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer), !tbaa [[TBAA11]]
// CHECK-NEXT:    ret <vscale x 8 x half> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f16u10__SVBool_tPKDhl(
// 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:%.*]] = bitcast half* [[BASE:%.*]] to <vscale x 8 x half>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 8 x half>, <vscale x 8 x half>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast half* [[TMP2]] to <vscale x 8 x half>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0nxv8f16(<vscale x 8 x half>* [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer), !tbaa [[TBAA11]]
// CPP-CHECK-NEXT:    ret <vscale x 8 x half> [[TMP4]]
//
svfloat16_t test_svld1_vnum_f16(svbool_t pg, const float16_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_f16,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_f32(
// 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:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x float>, <vscale x 4 x float>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast float* [[TMP2]] to <vscale x 4 x float>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0nxv4f32(<vscale x 4 x float>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer), !tbaa [[TBAA13]]
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f32u10__SVBool_tPKfl(
// 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:%.*]] = bitcast float* [[BASE:%.*]] to <vscale x 4 x float>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 4 x float>, <vscale x 4 x float>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast float* [[TMP2]] to <vscale x 4 x float>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0nxv4f32(<vscale x 4 x float>* [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer), !tbaa [[TBAA13]]
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP4]]
//
svfloat32_t test_svld1_vnum_f32(svbool_t pg, const float32_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_f32,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_vnum_f64(
// 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:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
// CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x double>, <vscale x 2 x double>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CHECK-NEXT:    [[TMP3:%.*]] = bitcast double* [[TMP2]] to <vscale x 2 x double>*
// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0nxv2f64(<vscale x 2 x double>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer), !tbaa [[TBAA15]]
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP4]]
//
// CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f64u10__SVBool_tPKdl(
// 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:%.*]] = bitcast double* [[BASE:%.*]] to <vscale x 2 x double>*
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = getelementptr <vscale x 2 x double>, <vscale x 2 x double>* [[TMP1]], i64 [[VNUM:%.*]], i64 0
// CPP-CHECK-NEXT:    [[TMP3:%.*]] = bitcast double* [[TMP2]] to <vscale x 2 x double>*
// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0nxv2f64(<vscale x 2 x double>* [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer), !tbaa [[TBAA15]]
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP4]]
//
svfloat64_t test_svld1_vnum_f64(svbool_t pg, const float64_t *base, int64_t vnum)
{
  return SVE_ACLE_FUNC(svld1_vnum,_f64,,)(pg, base, vnum);
}

// CHECK-LABEL: @test_svld1_gather_u32base_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u32base_s32u10__SVBool_tu12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_u32base_s32(svbool_t pg, svuint32_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _s32, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_u64base_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u64base_s64u10__SVBool_tu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_u64base_s64(svbool_t pg, svuint64_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _s64, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_u32base_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u32base_u32u10__SVBool_tu12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_u32base_u32(svbool_t pg, svuint32_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _u32, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_u64base_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u64base_u64u10__SVBool_tu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_u64base_u64(svbool_t pg, svuint64_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _u64, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_u32base_f32(
// 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.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u32base_f32u10__SVBool_tu12__SVUint32_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.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_u32base_f32(svbool_t pg, svuint32_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _f32, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_u64base_f64(
// 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.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z29test_svld1_gather_u64base_f64u10__SVBool_tu12__SVUint64_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.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_u64base_f64(svbool_t pg, svuint64_t bases) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _f64, )(pg, bases);
}

// CHECK-LABEL: @test_svld1_gather_s32offset_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s32offset_s32u10__SVBool_tPKiu11__SVInt32_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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_s32offset_s32(svbool_t pg, const int32_t *base, svint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, offset, _s32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_s64offset_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s64offset_s64u10__SVBool_tPKlu11__SVInt64_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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_s64offset_s64(svbool_t pg, const int64_t *base, svint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, offset, _s64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_s32offset_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s32offset_u32u10__SVBool_tPKju11__SVInt32_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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_s32offset_u32(svbool_t pg, const uint32_t *base, svint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, offset, _u32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_s64offset_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s64offset_u64u10__SVBool_tPKmu11__SVInt64_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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_s64offset_u64(svbool_t pg, const uint64_t *base, svint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, offset, _u64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_s32offset_f32(
// 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.ld1.gather.sxtw.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s32offset_f32u10__SVBool_tPKfu11__SVInt32_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.ld1.gather.sxtw.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_s32offset_f32(svbool_t pg, const float32_t *base, svint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, offset, _f32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_s64offset_f64(
// 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.ld1.gather.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_s64offset_f64u10__SVBool_tPKdu11__SVInt64_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.ld1.gather.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_s64offset_f64(svbool_t pg, const float64_t *base, svint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, offset, _f64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u32offset_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u32offset_s32u10__SVBool_tPKiu12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_u32offset_s32(svbool_t pg, const int32_t *base, svuint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, offset, _s32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u64offset_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u64offset_s64u10__SVBool_tPKlu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_u64offset_s64(svbool_t pg, const int64_t *base, svuint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, offset, _s64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u32offset_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u32offset_u32u10__SVBool_tPKju12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_u32offset_u32(svbool_t pg, const uint32_t *base, svuint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, offset, _u32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u64offset_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u64offset_u64u10__SVBool_tPKmu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_u64offset_u64(svbool_t pg, const uint64_t *base, svuint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, offset, _u64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u32offset_f32(
// 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.ld1.gather.uxtw.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u32offset_f32u10__SVBool_tPKfu12__SVUint32_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.ld1.gather.uxtw.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_u32offset_f32(svbool_t pg, const float32_t *base, svuint32_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, offset, _f32)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u64offset_f64(
// 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.ld1.gather.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z31test_svld1_gather_u64offset_f64u10__SVBool_tPKdu12__SVUint64_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.ld1.gather.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_u64offset_f64(svbool_t pg, const float64_t *base, svuint64_t offsets) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, offset, _f64)(pg, base, offsets);
}

// CHECK-LABEL: @test_svld1_gather_u32base_offset_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u32base_offset_s32u10__SVBool_tu12__SVUint32_tl(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_u32base_offset_s32(svbool_t pg, svuint32_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _offset_s32, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_u64base_offset_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u64base_offset_s64u10__SVBool_tu12__SVUint64_tl(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_u64base_offset_s64(svbool_t pg, svuint64_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _offset_s64, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_u32base_offset_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u32base_offset_u32u10__SVBool_tu12__SVUint32_tl(
// 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 i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_u32base_offset_u32(svbool_t pg, svuint32_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _offset_u32, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_u64base_offset_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u64base_offset_u64u10__SVBool_tu12__SVUint64_tl(
// 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 i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_u64base_offset_u64(svbool_t pg, svuint64_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _offset_u64, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_u32base_offset_f32(
// 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.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u32base_offset_f32u10__SVBool_tu12__SVUint32_tl(
// 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.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_u32base_offset_f32(svbool_t pg, svuint32_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _offset_f32, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_u64base_offset_f64(
// 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.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z36test_svld1_gather_u64base_offset_f64u10__SVBool_tu12__SVUint64_tl(
// 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.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_u64base_offset_f64(svbool_t pg, svuint64_t bases, int64_t offset) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _offset_f64, )(pg, bases, offset);
}

// CHECK-LABEL: @test_svld1_gather_s32index_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s32index_s32u10__SVBool_tPKiu11__SVInt32_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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_s32index_s32(svbool_t pg, const int32_t *base, svint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, index, _s32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_s64index_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s64index_s64u10__SVBool_tPKlu11__SVInt64_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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_s64index_s64(svbool_t pg, const int64_t *base, svint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, index, _s64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_s32index_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s32index_u32u10__SVBool_tPKju11__SVInt32_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 i32> @llvm.aarch64.sve.ld1.gather.sxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_s32index_u32(svbool_t pg, const uint32_t *base, svint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, index, _u32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_s64index_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s64index_u64u10__SVBool_tPKmu11__SVInt64_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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_s64index_u64(svbool_t pg, const uint64_t *base, svint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, index, _u64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_s32index_f32(
// 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.ld1.gather.sxtw.index.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s32index_f32u10__SVBool_tPKfu11__SVInt32_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.ld1.gather.sxtw.index.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_s32index_f32(svbool_t pg, const float32_t *base, svint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s32, index, _f32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_s64index_f64(
// 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.ld1.gather.index.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_s64index_f64u10__SVBool_tPKdu11__SVInt64_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.ld1.gather.index.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_s64index_f64(svbool_t pg, const float64_t *base, svint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, s64, index, _f64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u32index_s32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u32index_s32u10__SVBool_tPKiu12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svint32_t test_svld1_gather_u32index_s32(svbool_t pg, const int32_t *base, svuint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, index, _s32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u64index_s64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u64index_s64u10__SVBool_tPKlu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svint64_t test_svld1_gather_u64index_s64(svbool_t pg, const int64_t *base, svuint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, index, _s64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u32index_u32(
// 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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u32index_u32u10__SVBool_tPKju12__SVUint32_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 i32> @llvm.aarch64.sve.ld1.gather.uxtw.index.nxv4i32(<vscale x 4 x i1> [[TMP0]], i32* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP1]]
//
svuint32_t test_svld1_gather_u32index_u32(svbool_t pg, const uint32_t *base, svuint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, index, _u32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u64index_u64(
// 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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u64index_u64u10__SVBool_tPKmu12__SVUint64_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 i64> @llvm.aarch64.sve.ld1.gather.index.nxv2i64(<vscale x 2 x i1> [[TMP0]], i64* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP1]]
//
svuint64_t test_svld1_gather_u64index_u64(svbool_t pg, const uint64_t *base, svuint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, index, _u64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u32index_f32(
// 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.ld1.gather.uxtw.index.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u32index_f32u10__SVBool_tPKfu12__SVUint32_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.ld1.gather.uxtw.index.nxv4f32(<vscale x 4 x i1> [[TMP0]], float* [[BASE:%.*]], <vscale x 4 x i32> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP1]]
//
svfloat32_t test_svld1_gather_u32index_f32(svbool_t pg, const float32_t *base, svuint32_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u32, index, _f32)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u64index_f64(
// 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.ld1.gather.index.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
// CPP-CHECK-LABEL: @_Z30test_svld1_gather_u64index_f64u10__SVBool_tPKdu12__SVUint64_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.ld1.gather.index.nxv2f64(<vscale x 2 x i1> [[TMP0]], double* [[BASE:%.*]], <vscale x 2 x i64> [[INDICES:%.*]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP1]]
//
svfloat64_t test_svld1_gather_u64index_f64(svbool_t pg, const float64_t *base, svuint64_t indices) {
  return SVE_ACLE_FUNC(svld1_gather_, u64, index, _f64)(pg, base, indices);
}

// CHECK-LABEL: @test_svld1_gather_u32base_index_s32(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u32base_index_s32u10__SVBool_tu12__SVUint32_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
svint32_t test_svld1_gather_u32base_index_s32(svbool_t pg, svuint32_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _index_s32, )(pg, bases, index);
}

// CHECK-LABEL: @test_svld1_gather_u64base_index_s64(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u64base_index_s64u10__SVBool_tu12__SVUint64_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
svint64_t test_svld1_gather_u64base_index_s64(svbool_t pg, svuint64_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _index_s64, )(pg, bases, index);
}

// CHECK-LABEL: @test_svld1_gather_u32base_index_u32(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u32base_index_u32u10__SVBool_tu12__SVUint32_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x i32> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4i32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x i32> [[TMP2]]
//
svuint32_t test_svld1_gather_u32base_index_u32(svbool_t pg, svuint32_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _index_u32, )(pg, bases, index);
}

// CHECK-LABEL: @test_svld1_gather_u64base_index_u64(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u64base_index_u64u10__SVBool_tu12__SVUint64_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x i64> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2i64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x i64> [[TMP2]]
//
svuint64_t test_svld1_gather_u64base_index_u64(svbool_t pg, svuint64_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _index_u64, )(pg, bases, index);
}

// CHECK-LABEL: @test_svld1_gather_u32base_index_f32(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 4 x float> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u32base_index_f32u10__SVBool_tu12__SVUint32_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 2
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 4 x float> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv4f32.nxv4i32(<vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 4 x float> [[TMP2]]
//
svfloat32_t test_svld1_gather_u32base_index_f32(svbool_t pg, svuint32_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u32base, _index_f32, )(pg, bases, index);
}

// CHECK-LABEL: @test_svld1_gather_u64base_index_f64(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CHECK-NEXT:    ret <vscale x 2 x double> [[TMP2]]
//
// CPP-CHECK-LABEL: @_Z35test_svld1_gather_u64base_index_f64u10__SVBool_tu12__SVUint64_tl(
// 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:%.*]] = shl i64 [[INDEX:%.*]], 3
// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 2 x double> @llvm.aarch64.sve.ld1.gather.scalar.offset.nxv2f64.nxv2i64(<vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[TMP1]])
// CPP-CHECK-NEXT:    ret <vscale x 2 x double> [[TMP2]]
//
svfloat64_t test_svld1_gather_u64base_index_f64(svbool_t pg, svuint64_t bases, int64_t index) {
  return SVE_ACLE_FUNC(svld1_gather, _u64base, _index_f64, )(pg, bases, index);
}