// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --include-generated-funcs --version 5
// RUN: %clang_cc1 %s -emit-llvm -o - -cl-std=CL2.0 -O0 -triple amdgcn | FileCheck -check-prefix=AMDGCN %s

typedef int int2 __attribute__((ext_vector_type(2)));

typedef struct {
  int cells[9];
} Mat3X3;

typedef struct {
  int cells[16];
} Mat4X4;

typedef struct {
  int cells[1024];
} Mat32X32;

typedef struct {
  int cells[4096];
} Mat64X64;

struct StructOneMember {
  int2 x;
};

struct StructTwoMember {
  int2 x;
  int2 y;
};

struct LargeStructOneMember {
  int2 x[100];
};

struct LargeStructTwoMember {
  int2 x[40];
  int2 y[20];
};

#if (__OPENCL_C_VERSION__ == 200) || (__OPENCL_C_VERSION__ >= 300 && defined(__opencl_c_program_scope_global_variables))
struct LargeStructOneMember g_s;
#endif


Mat4X4 __attribute__((noinline)) foo(Mat3X3 in) {
  Mat4X4 out;
  return out;
}

// Expect two mem copies: one for the argument "in", and one for
// the return value.

kernel void ker(global Mat3X3 *in, global Mat4X4 *out) {
  out[0] = foo(in[1]);
}

Mat64X64 __attribute__((noinline)) foo_large(Mat32X32 in) {
  Mat64X64 out;
  return out;
}

kernel void ker_large(global Mat32X32 *in, global Mat64X64 *out) {
  out[0] = foo_large(in[1]);
}

void FuncOneMember(struct StructOneMember u) {
  u.x = (int2)(0, 0);
}

void FuncOneLargeMember(struct LargeStructOneMember u) {
  u.x[0] = (int2)(0, 0);
}

#if (__OPENCL_C_VERSION__ == 200) || (__OPENCL_C_VERSION__ >= 300 && defined(__opencl_c_program_scope_global_variables))
void test_indirect_arg_globl(void) {
  FuncOneLargeMember(g_s);
}
#endif

kernel void test_indirect_arg_local(void) {
  local struct LargeStructOneMember l_s;
  FuncOneLargeMember(l_s);
}

void test_indirect_arg_private(void) {
  struct LargeStructOneMember p_s;
  FuncOneLargeMember(p_s);
}

kernel void KernelOneMember(struct StructOneMember u) {
  FuncOneMember(u);
}

kernel void KernelOneMemberSpir(global struct StructOneMember* u) {
  FuncOneMember(*u);
}

kernel void KernelLargeOneMember(struct LargeStructOneMember u) {
  FuncOneLargeMember(u);
}

void FuncTwoMember(struct StructTwoMember u) {
  u.y = (int2)(0, 0);
}

void FuncLargeTwoMember(struct LargeStructTwoMember u) {
  u.y[0] = (int2)(0, 0);
}

kernel void KernelTwoMember(struct StructTwoMember u) {
  FuncTwoMember(u);
}

kernel void KernelLargeTwoMember(struct LargeStructTwoMember u) {
  FuncLargeTwoMember(u);
}
// AMDGCN-LABEL: define dso_local %struct.Mat4X4 @foo(
// AMDGCN-SAME: [9 x i32] [[IN_COERCE:%.*]]) #[[ATTR0:[0-9]+]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[RETVAL:%.*]] = alloca [[STRUCT_MAT4X4:%.*]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[IN:%.*]] = alloca [[STRUCT_MAT3X3:%.*]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[RETVAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[RETVAL]] to ptr
// AMDGCN-NEXT:    [[IN1:%.*]] = addrspacecast ptr addrspace(5) [[IN]] to ptr
// AMDGCN-NEXT:    [[COERCE_DIVE:%.*]] = getelementptr inbounds nuw [[STRUCT_MAT3X3]], ptr [[IN1]], i32 0, i32 0
// AMDGCN-NEXT:    store [9 x i32] [[IN_COERCE]], ptr [[COERCE_DIVE]], align 4
// AMDGCN-NEXT:    [[TMP0:%.*]] = load [[STRUCT_MAT4X4]], ptr [[RETVAL_ASCAST]], align 4
// AMDGCN-NEXT:    ret [[STRUCT_MAT4X4]] [[TMP0]]
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @ker(
// AMDGCN-SAME: ptr addrspace(1) noundef align 4 [[IN:%.*]], ptr addrspace(1) noundef align 4 [[OUT:%.*]]) #[[ATTR1:[0-9]+]] !kernel_arg_addr_space [[META4:![0-9]+]] !kernel_arg_access_qual [[META5:![0-9]+]] !kernel_arg_type [[META6:![0-9]+]] !kernel_arg_base_type [[META6]] !kernel_arg_type_qual [[META7:![0-9]+]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[IN_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[OUT_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[IN_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[IN_ADDR]] to ptr
// AMDGCN-NEXT:    [[OUT_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[OUT_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[IN]], ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    store ptr addrspace(1) [[OUT]], ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = load ptr addrspace(1), ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_ker(ptr addrspace(1) noundef align 4 [[TMP0]], ptr addrspace(1) noundef align 4 [[TMP1]]) #[[ATTR4:[0-9]+]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_ker(
// AMDGCN-SAME: ptr addrspace(1) noundef align 4 [[IN:%.*]], ptr addrspace(1) noundef align 4 [[OUT:%.*]]) #[[ATTR2:[0-9]+]] !kernel_arg_addr_space [[META4]] !kernel_arg_access_qual [[META5]] !kernel_arg_type [[META6]] !kernel_arg_base_type [[META6]] !kernel_arg_type_qual [[META7]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[IN_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[OUT_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[TMP:%.*]] = alloca [[STRUCT_MAT4X4:%.*]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[IN_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[IN_ADDR]] to ptr
// AMDGCN-NEXT:    [[OUT_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[OUT_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[IN]], ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    store ptr addrspace(1) [[OUT]], ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [[STRUCT_MAT4X4]], ptr addrspace(1) [[TMP0]], i64 0
// AMDGCN-NEXT:    [[TMP1:%.*]] = load ptr addrspace(1), ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_MAT3X3:%.*]], ptr addrspace(1) [[TMP1]], i64 1
// AMDGCN-NEXT:    [[TMP2:%.*]] = getelementptr inbounds nuw [[STRUCT_MAT3X3]], ptr addrspace(1) [[ARRAYIDX1]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP3:%.*]] = load [9 x i32], ptr addrspace(1) [[TMP2]], align 4
// AMDGCN-NEXT:    [[CALL:%.*]] = call [[STRUCT_MAT4X4]] @[[FOO:[a-zA-Z0-9_$\"\\.-]*[a-zA-Z_$\"\\.-][a-zA-Z0-9_$\"\\.-]*]]([9 x i32] [[TMP3]]) #[[ATTR4]]
// AMDGCN-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_MAT4X4]], ptr addrspace(5) [[TMP]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP5:%.*]] = extractvalue [[STRUCT_MAT4X4]] [[CALL]], 0
// AMDGCN-NEXT:    store [16 x i32] [[TMP5]], ptr addrspace(5) [[TMP4]], align 4
// AMDGCN-NEXT:    call void @llvm.memcpy.p1.p5.i64(ptr addrspace(1) align 4 [[ARRAYIDX]], ptr addrspace(5) align 4 [[TMP]], i64 64, i1 false)
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @foo_large(
// AMDGCN-SAME: ptr addrspace(5) dead_on_unwind noalias writable sret([[STRUCT_MAT64X64:%.*]]) align 4 [[AGG_RESULT:%.*]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32:%.*]]) align 4 [[TMP0:%.*]]) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_MAT32X32]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[IN:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p5.i64(ptr align 4 [[IN]], ptr addrspace(5) align 4 [[TMP0]], i64 4096, i1 false)
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @ker_large(
// AMDGCN-SAME: ptr addrspace(1) noundef align 4 [[IN:%.*]], ptr addrspace(1) noundef align 4 [[OUT:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META4]] !kernel_arg_access_qual [[META5]] !kernel_arg_type [[META8:![0-9]+]] !kernel_arg_base_type [[META8]] !kernel_arg_type_qual [[META7]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[IN_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[OUT_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[IN_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[IN_ADDR]] to ptr
// AMDGCN-NEXT:    [[OUT_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[OUT_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[IN]], ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    store ptr addrspace(1) [[OUT]], ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = load ptr addrspace(1), ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_ker_large(ptr addrspace(1) noundef align 4 [[TMP0]], ptr addrspace(1) noundef align 4 [[TMP1]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_ker_large(
// AMDGCN-SAME: ptr addrspace(1) noundef align 4 [[IN:%.*]], ptr addrspace(1) noundef align 4 [[OUT:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META4]] !kernel_arg_access_qual [[META5]] !kernel_arg_type [[META8]] !kernel_arg_base_type [[META8]] !kernel_arg_type_qual [[META7]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[IN_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[OUT_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[TMP:%.*]] = alloca [[STRUCT_MAT64X64:%.*]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[BYVAL_TEMP:%.*]] = alloca [[STRUCT_MAT32X32:%.*]], align 4, addrspace(5)
// AMDGCN-NEXT:    [[IN_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[IN_ADDR]] to ptr
// AMDGCN-NEXT:    [[OUT_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[OUT_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[IN]], ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    store ptr addrspace(1) [[OUT]], ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[OUT_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [[STRUCT_MAT64X64]], ptr addrspace(1) [[TMP0]], i64 0
// AMDGCN-NEXT:    [[TMP1:%.*]] = load ptr addrspace(1), ptr [[IN_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[ARRAYIDX1:%.*]] = getelementptr inbounds [[STRUCT_MAT32X32]], ptr addrspace(1) [[TMP1]], i64 1
// AMDGCN-NEXT:    call void @llvm.memcpy.p5.p1.i64(ptr addrspace(5) align 4 [[BYVAL_TEMP]], ptr addrspace(1) align 4 [[ARRAYIDX1]], i64 4096, i1 false)
// AMDGCN-NEXT:    call void @foo_large(ptr addrspace(5) dead_on_unwind writable sret([[STRUCT_MAT64X64]]) align 4 [[TMP]], ptr addrspace(5) noundef byref([[STRUCT_MAT32X32]]) align 4 [[BYVAL_TEMP]]) #[[ATTR4]]
// AMDGCN-NEXT:    call void @llvm.memcpy.p1.p5.i64(ptr addrspace(1) align 4 [[ARRAYIDX]], ptr addrspace(5) align 4 [[TMP]], i64 16384, i1 false)
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @FuncOneMember(
// AMDGCN-SAME: <2 x i32> [[U_COERCE:%.*]]) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U:%.*]] = alloca [[STRUCT_STRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL:%.*]] = alloca <2 x i32>, align 8, addrspace(5)
// AMDGCN-NEXT:    [[U1:%.*]] = addrspacecast ptr addrspace(5) [[U]] to ptr
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[DOTCOMPOUNDLITERAL]] to ptr
// AMDGCN-NEXT:    [[COERCE_DIVE:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE]], ptr [[COERCE_DIVE]], align 8
// AMDGCN-NEXT:    store <2 x i32> zeroinitializer, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load <2 x i32>, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[X:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[TMP0]], ptr [[X]], align 8
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @FuncOneLargeMember(
// AMDGCN-SAME: ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL:%.*]] = alloca <2 x i32>, align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[DOTCOMPOUNDLITERAL]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p5.i64(ptr align 8 [[U]], ptr addrspace(5) align 8 [[TMP0]], i64 800, i1 false)
// AMDGCN-NEXT:    store <2 x i32> zeroinitializer, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = load <2 x i32>, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[X:%.*]] = getelementptr inbounds nuw [[STRUCT_LARGESTRUCTONEMEMBER]], ptr [[U]], i32 0, i32 0
// AMDGCN-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [100 x <2 x i32>], ptr [[X]], i64 0, i64 0
// AMDGCN-NEXT:    store <2 x i32> [[TMP1]], ptr [[ARRAYIDX]], align 8
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @test_indirect_arg_globl(
// AMDGCN-SAME: ) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[BYVAL_TEMP:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    call void @llvm.memcpy.p5.p1.i64(ptr addrspace(5) align 8 [[BYVAL_TEMP]], ptr addrspace(1) align 8 @g_s, i64 800, i1 false)
// AMDGCN-NEXT:    call void @FuncOneLargeMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER]]) align 8 [[BYVAL_TEMP]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @test_indirect_arg_local(
// AMDGCN-SAME: ) #[[ATTR1]] !kernel_arg_addr_space [[META9:![0-9]+]] !kernel_arg_access_qual [[META9]] !kernel_arg_type [[META9]] !kernel_arg_base_type [[META9]] !kernel_arg_type_qual [[META9]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_test_indirect_arg_local() #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_test_indirect_arg_local(
// AMDGCN-SAME: ) #[[ATTR2]] !kernel_arg_addr_space [[META9]] !kernel_arg_access_qual [[META9]] !kernel_arg_type [[META9]] !kernel_arg_base_type [[META9]] !kernel_arg_type_qual [[META9]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[BYVAL_TEMP:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    call void @llvm.memcpy.p5.p3.i64(ptr addrspace(5) align 8 [[BYVAL_TEMP]], ptr addrspace(3) align 8 @test_indirect_arg_local.l_s, i64 800, i1 false)
// AMDGCN-NEXT:    call void @FuncOneLargeMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER]]) align 8 [[BYVAL_TEMP]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @test_indirect_arg_private(
// AMDGCN-SAME: ) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[P_S:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    call void @FuncOneLargeMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER]]) align 8 [[P_S]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @KernelOneMember(
// AMDGCN-SAME: <2 x i32> [[U_COERCE:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META10:![0-9]+]] !kernel_arg_access_qual [[META11:![0-9]+]] !kernel_arg_type [[META12:![0-9]+]] !kernel_arg_base_type [[META12]] !kernel_arg_type_qual [[META13:![0-9]+]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U:%.*]] = alloca [[STRUCT_STRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U1:%.*]] = addrspacecast ptr addrspace(5) [[U]] to ptr
// AMDGCN-NEXT:    [[COERCE_DIVE:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE]], ptr [[COERCE_DIVE]], align 8
// AMDGCN-NEXT:    [[COERCE_DIVE2:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP0:%.*]] = load <2 x i32>, ptr [[COERCE_DIVE2]], align 8
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_KernelOneMember(<2 x i32> [[TMP0]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_KernelOneMember(
// AMDGCN-SAME: <2 x i32> [[U_COERCE:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META12]] !kernel_arg_base_type [[META12]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U:%.*]] = alloca [[STRUCT_STRUCTONEMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U1:%.*]] = addrspacecast ptr addrspace(5) [[U]] to ptr
// AMDGCN-NEXT:    [[COERCE_DIVE:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE]], ptr [[COERCE_DIVE]], align 8
// AMDGCN-NEXT:    [[COERCE_DIVE2:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP0:%.*]] = load <2 x i32>, ptr [[COERCE_DIVE2]], align 8
// AMDGCN-NEXT:    call void @FuncOneMember(<2 x i32> [[TMP0]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @KernelOneMemberSpir(
// AMDGCN-SAME: ptr addrspace(1) noundef align 8 [[U:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META14:![0-9]+]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META15:![0-9]+]] !kernel_arg_base_type [[META15]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[U_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[U_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[U]], ptr [[U_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[U_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_KernelOneMemberSpir(ptr addrspace(1) noundef align 8 [[TMP0]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_KernelOneMemberSpir(
// AMDGCN-SAME: ptr addrspace(1) noundef align 8 [[U:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META14]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META15]] !kernel_arg_base_type [[META15]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U_ADDR:%.*]] = alloca ptr addrspace(1), align 8, addrspace(5)
// AMDGCN-NEXT:    [[U_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[U_ADDR]] to ptr
// AMDGCN-NEXT:    store ptr addrspace(1) [[U]], ptr [[U_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP0:%.*]] = load ptr addrspace(1), ptr [[U_ADDR_ASCAST]], align 8
// AMDGCN-NEXT:    [[COERCE_DIVE:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTONEMEMBER:%.*]], ptr addrspace(1) [[TMP0]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP1:%.*]] = load <2 x i32>, ptr addrspace(1) [[COERCE_DIVE]], align 8
// AMDGCN-NEXT:    call void @FuncOneMember(<2 x i32> [[TMP1]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @KernelLargeOneMember(
// AMDGCN-SAME: ptr addrspace(4) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META16:![0-9]+]] !kernel_arg_base_type [[META16]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p4.i64(ptr align 8 [[U]], ptr addrspace(4) align 8 [[TMP0]], i64 800, i1 false)
// AMDGCN-NEXT:    [[U_ASCAST:%.*]] = addrspacecast ptr [[U]] to ptr addrspace(5)
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_KernelLargeOneMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER]]) align 8 [[U_ASCAST]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_KernelLargeOneMember(
// AMDGCN-SAME: ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META16]] !kernel_arg_base_type [[META16]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[BYVAL_TEMP:%.*]] = alloca [[STRUCT_LARGESTRUCTONEMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p5.i64(ptr align 8 [[U]], ptr addrspace(5) align 8 [[TMP0]], i64 800, i1 false)
// AMDGCN-NEXT:    call void @llvm.memcpy.p5.p0.i64(ptr addrspace(5) align 8 [[BYVAL_TEMP]], ptr align 8 [[U]], i64 800, i1 false)
// AMDGCN-NEXT:    call void @FuncOneLargeMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTONEMEMBER]]) align 8 [[BYVAL_TEMP]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @FuncTwoMember(
// AMDGCN-SAME: <2 x i32> [[U_COERCE0:%.*]], <2 x i32> [[U_COERCE1:%.*]]) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U:%.*]] = alloca [[STRUCT_STRUCTTWOMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL:%.*]] = alloca <2 x i32>, align 8, addrspace(5)
// AMDGCN-NEXT:    [[U1:%.*]] = addrspacecast ptr addrspace(5) [[U]] to ptr
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[DOTCOMPOUNDLITERAL]] to ptr
// AMDGCN-NEXT:    [[TMP0:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE0]], ptr [[TMP0]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 1
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE1]], ptr [[TMP1]], align 8
// AMDGCN-NEXT:    store <2 x i32> zeroinitializer, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP2:%.*]] = load <2 x i32>, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[Y:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 1
// AMDGCN-NEXT:    store <2 x i32> [[TMP2]], ptr [[Y]], align 8
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @FuncLargeTwoMember(
// AMDGCN-SAME: ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTTWOMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR0]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTTWOMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL:%.*]] = alloca <2 x i32>, align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    [[DOTCOMPOUNDLITERAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[DOTCOMPOUNDLITERAL]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p5.i64(ptr align 8 [[U]], ptr addrspace(5) align 8 [[TMP0]], i64 480, i1 false)
// AMDGCN-NEXT:    store <2 x i32> zeroinitializer, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = load <2 x i32>, ptr [[DOTCOMPOUNDLITERAL_ASCAST]], align 8
// AMDGCN-NEXT:    [[Y:%.*]] = getelementptr inbounds nuw [[STRUCT_LARGESTRUCTTWOMEMBER]], ptr [[U]], i32 0, i32 1
// AMDGCN-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [20 x <2 x i32>], ptr [[Y]], i64 0, i64 0
// AMDGCN-NEXT:    store <2 x i32> [[TMP1]], ptr [[ARRAYIDX]], align 8
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @KernelTwoMember(
// AMDGCN-SAME: ptr addrspace(4) noundef byref([[STRUCT_STRUCTTWOMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META17:![0-9]+]] !kernel_arg_base_type [[META17]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_STRUCTTWOMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p4.i64(ptr align 8 [[U]], ptr addrspace(4) align 8 [[TMP0]], i64 16, i1 false)
// AMDGCN-NEXT:    [[TMP1:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP2:%.*]] = load <2 x i32>, ptr [[TMP1]], align 8
// AMDGCN-NEXT:    [[TMP3:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U]], i32 0, i32 1
// AMDGCN-NEXT:    [[TMP4:%.*]] = load <2 x i32>, ptr [[TMP3]], align 8
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_KernelTwoMember(<2 x i32> [[TMP2]], <2 x i32> [[TMP4]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_KernelTwoMember(
// AMDGCN-SAME: <2 x i32> [[U_COERCE0:%.*]], <2 x i32> [[U_COERCE1:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META17]] !kernel_arg_base_type [[META17]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[U:%.*]] = alloca [[STRUCT_STRUCTTWOMEMBER:%.*]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U1:%.*]] = addrspacecast ptr addrspace(5) [[U]] to ptr
// AMDGCN-NEXT:    [[TMP0:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE0]], ptr [[TMP0]], align 8
// AMDGCN-NEXT:    [[TMP1:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 1
// AMDGCN-NEXT:    store <2 x i32> [[U_COERCE1]], ptr [[TMP1]], align 8
// AMDGCN-NEXT:    [[TMP2:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 0
// AMDGCN-NEXT:    [[TMP3:%.*]] = load <2 x i32>, ptr [[TMP2]], align 8
// AMDGCN-NEXT:    [[TMP4:%.*]] = getelementptr inbounds nuw [[STRUCT_STRUCTTWOMEMBER]], ptr [[U1]], i32 0, i32 1
// AMDGCN-NEXT:    [[TMP5:%.*]] = load <2 x i32>, ptr [[TMP4]], align 8
// AMDGCN-NEXT:    call void @FuncTwoMember(<2 x i32> [[TMP3]], <2 x i32> [[TMP5]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local amdgpu_kernel void @KernelLargeTwoMember(
// AMDGCN-SAME: ptr addrspace(4) noundef byref([[STRUCT_LARGESTRUCTTWOMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR1]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META18:![0-9]+]] !kernel_arg_base_type [[META18]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTTWOMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p4.i64(ptr align 8 [[U]], ptr addrspace(4) align 8 [[TMP0]], i64 480, i1 false)
// AMDGCN-NEXT:    [[U_ASCAST:%.*]] = addrspacecast ptr [[U]] to ptr addrspace(5)
// AMDGCN-NEXT:    call void @__clang_ocl_kern_imp_KernelLargeTwoMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTTWOMEMBER]]) align 8 [[U_ASCAST]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//
// AMDGCN-LABEL: define dso_local void @__clang_ocl_kern_imp_KernelLargeTwoMember(
// AMDGCN-SAME: ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTTWOMEMBER:%.*]]) align 8 [[TMP0:%.*]]) #[[ATTR2]] !kernel_arg_addr_space [[META10]] !kernel_arg_access_qual [[META11]] !kernel_arg_type [[META18]] !kernel_arg_base_type [[META18]] !kernel_arg_type_qual [[META13]] {
// AMDGCN-NEXT:  [[ENTRY:.*:]]
// AMDGCN-NEXT:    [[COERCE:%.*]] = alloca [[STRUCT_LARGESTRUCTTWOMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[BYVAL_TEMP:%.*]] = alloca [[STRUCT_LARGESTRUCTTWOMEMBER]], align 8, addrspace(5)
// AMDGCN-NEXT:    [[U:%.*]] = addrspacecast ptr addrspace(5) [[COERCE]] to ptr
// AMDGCN-NEXT:    call void @llvm.memcpy.p0.p5.i64(ptr align 8 [[U]], ptr addrspace(5) align 8 [[TMP0]], i64 480, i1 false)
// AMDGCN-NEXT:    call void @llvm.memcpy.p5.p0.i64(ptr addrspace(5) align 8 [[BYVAL_TEMP]], ptr align 8 [[U]], i64 480, i1 false)
// AMDGCN-NEXT:    call void @FuncLargeTwoMember(ptr addrspace(5) noundef byref([[STRUCT_LARGESTRUCTTWOMEMBER]]) align 8 [[BYVAL_TEMP]]) #[[ATTR4]]
// AMDGCN-NEXT:    ret void
//
//.
// AMDGCN: [[META4]] = !{i32 1, i32 1}
// AMDGCN: [[META5]] = !{!"none", !"none"}
// AMDGCN: [[META6]] = !{!"Mat3X3*", !"Mat4X4*"}
// AMDGCN: [[META7]] = !{!"", !""}
// AMDGCN: [[META8]] = !{!"Mat32X32*", !"Mat64X64*"}
// AMDGCN: [[META9]] = !{}
// AMDGCN: [[META10]] = !{i32 0}
// AMDGCN: [[META11]] = !{!"none"}
// AMDGCN: [[META12]] = !{!"struct StructOneMember"}
// AMDGCN: [[META13]] = !{!""}
// AMDGCN: [[META14]] = !{i32 1}
// AMDGCN: [[META15]] = !{!"struct StructOneMember*"}
// AMDGCN: [[META16]] = !{!"struct LargeStructOneMember"}
// AMDGCN: [[META17]] = !{!"struct StructTwoMember"}
// AMDGCN: [[META18]] = !{!"struct LargeStructTwoMember"}
//.