llvm/clang/test/CodeGen/aarch64-sve-intrinsics/acle_sve_st1b.c

// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: aarch64-registered-target
// RUN: %clang_cc1 -triple aarch64 -target-feature +sve -disable-O0-optnone -Werror -o - -emit-llvm %s | opt -S -passes=mem2reg,instcombine,tailcallelim | FileCheck %s
// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64 -target-feature +sve -disable-O0-optnone -Werror -o - -emit-llvm %s | opt -S -passes=mem2reg,instcombine,tailcallelim | FileCheck %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sve -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
// RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -S -disable-O0-optnone -Werror -Wall -o /dev/null %s

#include <arm_sve.h>

#if defined __ARM_FEATURE_SME
#define MODE_ATTR __arm_streaming
#else
#define MODE_ATTR
#endif

#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_svst1b_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:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_s16(svbool_t pg, int8_t *base, svint16_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_s16,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_s32(svbool_t pg, int8_t *base, svint32_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_s32,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_s64(svbool_t pg, int8_t *base, svint64_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_s64,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_u16(svbool_t pg, uint8_t *base, svuint16_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_u16,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_u32(svbool_t pg, uint8_t *base, svuint32_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_u32,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_u64(svbool_t pg, uint8_t *base, svuint64_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b,_u64,,)(pg, base, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_s16(svbool_t pg, int8_t *base, int64_t vnum, svint16_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_s16,,)(pg, base, vnum, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_s32(svbool_t pg, int8_t *base, int64_t vnum, svint32_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_s32,,)(pg, base, vnum, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_s64(svbool_t pg, int8_t *base, int64_t vnum, svint64_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_s64,,)(pg, base, vnum, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_u16(svbool_t pg, uint8_t *base, int64_t vnum, svuint16_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_u16,,)(pg, base, vnum, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_u32(svbool_t pg, uint8_t *base, int64_t vnum, svuint32_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_u32,,)(pg, base, vnum, data);
}

// CHECK-LABEL: @test_svst1b_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:%.*]] = tail call i64 @llvm.vscale.i64()
// CHECK-NEXT:    [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1
// CHECK-NEXT:    [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]]
// CHECK-NEXT:    [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]]
// CHECK-NEXT:    [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]])
// CHECK-NEXT:    ret void
//
void test_svst1b_vnum_u64(svbool_t pg, uint8_t *base, int64_t vnum, svuint64_t data) MODE_ATTR
{
  return SVE_ACLE_FUNC(svst1b_vnum,_u64,,)(pg, base, vnum, data);
}

#ifndef __ARM_FEATURE_SME

// CHECK-LABEL: @test_svst1b_scatter_u32base_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32base_s32(svbool_t pg, svuint32_t bases, svint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u32base,,_s32)(pg, bases, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64base_s64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv2i8.nxv2i64(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64base_s64(svbool_t pg, svuint64_t bases, svint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u64base,,_s64)(pg, bases, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u32base_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32base_u32(svbool_t pg, svuint32_t bases, svuint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u32base,,_u32)(pg, bases, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64base_u64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv2i8.nxv2i64(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], <vscale x 2 x i64> [[BASES:%.*]], i64 0)
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64base_u64(svbool_t pg, svuint64_t bases, svuint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u64base,,_u64)(pg, bases, data);
}

// CHECK-LABEL: @test_svst1b_scatter_s32offset_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.sxtw.nxv4i8(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_s32offset_s32(svbool_t pg, int8_t *base, svint32_t offsets, svint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,s32,offset,_s32)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_s64offset_s64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.nxv2i8(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_s64offset_s64(svbool_t pg, int8_t *base, svint64_t offsets, svint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,s64,offset,_s64)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_s32offset_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.sxtw.nxv4i8(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_s32offset_u32(svbool_t pg, uint8_t *base, svint32_t offsets, svuint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,s32,offset,_u32)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_s64offset_u64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.nxv2i8(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_s64offset_u64(svbool_t pg, uint8_t *base, svint64_t offsets, svuint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,s64,offset,_u64)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u32offset_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.uxtw.nxv4i8(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32offset_s32(svbool_t pg, int8_t *base, svuint32_t offsets, svint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,u32,offset,_s32)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64offset_s64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.nxv2i8(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64offset_s64(svbool_t pg, int8_t *base, svuint64_t offsets, svint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,u64,offset,_s64)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u32offset_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.uxtw.nxv4i8(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 4 x i32> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32offset_u32(svbool_t pg, uint8_t *base, svuint32_t offsets, svuint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,u32,offset,_u32)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64offset_u64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.nxv2i8(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], ptr [[BASE:%.*]], <vscale x 2 x i64> [[OFFSETS:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64offset_u64(svbool_t pg, uint8_t *base, svuint64_t offsets, svuint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter_,u64,offset,_u64)(pg, base, offsets, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u32base_offset_s32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32base_offset_s32(svbool_t pg, svuint32_t bases, int64_t offset, svint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u32base,_offset,_s32)(pg, bases, offset, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64base_offset_s64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv2i8.nxv2i64(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64base_offset_s64(svbool_t pg, svuint64_t bases, int64_t offset, svint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u64base,_offset,_s64)(pg, bases, offset, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u32base_offset_u32(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv4i8.nxv4i32(<vscale x 4 x i8> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u32base_offset_u32(svbool_t pg, svuint32_t bases, int64_t offset, svuint32_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u32base,_offset,_u32)(pg, bases, offset, data);
}

// CHECK-LABEL: @test_svst1b_scatter_u64base_offset_u64(
// CHECK-NEXT:  entry:
// CHECK-NEXT:    [[TMP0:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8>
// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]])
// CHECK-NEXT:    tail call void @llvm.aarch64.sve.st1.scatter.scalar.offset.nxv2i8.nxv2i64(<vscale x 2 x i8> [[TMP0]], <vscale x 2 x i1> [[TMP1]], <vscale x 2 x i64> [[BASES:%.*]], i64 [[OFFSET:%.*]])
// CHECK-NEXT:    ret void
//
void test_svst1b_scatter_u64base_offset_u64(svbool_t pg, svuint64_t bases, int64_t offset, svuint64_t data)
{
  return SVE_ACLE_FUNC(svst1b_scatter,_u64base,_offset,_u64)(pg, bases, offset, data);
}

#endif