; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -mtriple=amdgcn-amd-amdhsa -S -amdgpu-aa-wrapper -amdgpu-aa -instcombine -o - %s | FileCheck %s

; Make sure the optimization from memcpy-from-global.ll happens, but
; the constant source is not a global variable.

target datalayout = "e-p:64:64-p1:64:64-p2:32:32-p3:32:32-p4:64:64-p5:32:32-p6:32:32-p7:160:256:256:32-p8:128:128-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-S32-A5"

; Simple memcpy to alloca from constant address space argument.
define i8 @memcpy_constant_arg_ptr_to_alloca(ptr addrspace(4) noalias readonly align 4 dereferenceable(32) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i8], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(4) [[GEP]], align 1
; CHECK-NEXT:    ret i8 [[LOAD]]
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  ret i8 %load
}

define i8 @memcpy_constant_arg_ptr_to_alloca_load_metadata(ptr addrspace(4) noalias readonly align 4 dereferenceable(32) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca_load_metadata(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i8], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(4) [[GEP]], align 1, !noalias !0
; CHECK-NEXT:    ret i8 [[LOAD]]
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep, !noalias !0
  ret i8 %load
}

define i64 @memcpy_constant_arg_ptr_to_alloca_load_alignment(ptr addrspace(4) noalias readonly align 4 dereferenceable(256) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca_load_alignment(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i64], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i64, ptr addrspace(4) [[GEP]], align 16
; CHECK-NEXT:    ret i64 [[LOAD]]
;
  %alloca = alloca [32 x i64], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 256, i1 false)
  %gep = getelementptr inbounds [32 x i64], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i64, ptr addrspace(5) %gep, align 16
  ret i64 %load
}

define i64 @memcpy_constant_arg_ptr_to_alloca_load_atomic(ptr addrspace(4) noalias readonly align 8 dereferenceable(256) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca_load_atomic(
; CHECK-NEXT:    [[ALLOCA:%.*]] = alloca [32 x i64], align 8, addrspace(5)
; CHECK-NEXT:    call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) noundef align 8 dereferenceable(256) [[ALLOCA]], ptr addrspace(4) noundef align 8 dereferenceable(256) [[ARG:%.*]], i64 256, i1 false)
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds [32 x i64], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[IDX:%.*]]
; CHECK-NEXT:    [[LOAD:%.*]] = load atomic i64, ptr addrspace(5) [[GEP]] syncscope("somescope") acquire, align 8
; CHECK-NEXT:    ret i64 [[LOAD]]
;
  %alloca = alloca [32 x i64], align 8, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 256, i1 false)
  %gep = getelementptr inbounds [32 x i64], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load atomic i64, ptr addrspace(5) %gep syncscope("somescope") acquire, align 8
  ret i64 %load
}

; Simple memmove to alloca from constant address space argument.
define i8 @memmove_constant_arg_ptr_to_alloca(ptr addrspace(4) noalias readonly align 4 dereferenceable(32) %arg, i32 %idx) {
; CHECK-LABEL: @memmove_constant_arg_ptr_to_alloca(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i8], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(4) [[GEP]], align 1
; CHECK-NEXT:    ret i8 [[LOAD]]
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  call void @llvm.memmove.p5.p4.i32(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i32 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  ret i8 %load
}

; Simple memcpy to alloca from byref constant address space argument.
define amdgpu_kernel void @memcpy_constant_byref_arg_ptr_to_alloca(ptr addrspace(4) noalias readonly align 4 byref([32 x i8]) %arg, ptr addrspace(1) %out, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_byref_arg_ptr_to_alloca(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i8], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(4) [[GEP]], align 1
; CHECK-NEXT:    store i8 [[LOAD]], ptr addrspace(1) [[OUT:%.*]], align 1
; CHECK-NEXT:    ret void
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  store i8 %load, ptr addrspace(1) %out
  ret void
}

; Simple memcpy to alloca from byref constant address space argument, but not enough bytes are dereferenceable
define amdgpu_kernel void @memcpy_constant_byref_arg_ptr_to_alloca_too_many_bytes(ptr addrspace(4) noalias readonly align 4 byref([31 x i8]) %arg, ptr addrspace(1) %out, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_byref_arg_ptr_to_alloca_too_many_bytes(
; CHECK-NEXT:    [[ALLOCA:%.*]] = alloca [32 x i8], align 4, addrspace(5)
; CHECK-NEXT:    call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) noundef align 4 dereferenceable(31) [[ALLOCA]], ptr addrspace(4) noundef align 4 dereferenceable(31) [[ARG:%.*]], i64 31, i1 false)
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds [32 x i8], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[IDX:%.*]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(5) [[GEP]], align 1
; CHECK-NEXT:    store i8 [[LOAD]], ptr addrspace(1) [[OUT:%.*]], align 1
; CHECK-NEXT:    ret void
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %arg, i64 31, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  store i8 %load, ptr addrspace(1) %out
  ret void
}

; Simple memcpy to alloca from constant address space intrinsic call
define amdgpu_kernel void @memcpy_constant_intrinsic_ptr_to_alloca(ptr addrspace(1) %out, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_intrinsic_ptr_to_alloca(
; CHECK-NEXT:    [[ALLOCA:%.*]] = alloca [32 x i8], align 4, addrspace(5)
; CHECK-NEXT:    [[KERNARG_SEGMENT_PTR:%.*]] = call align 16 dereferenceable(32) ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr()
; CHECK-NEXT:    call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) noundef align 4 dereferenceable(32) [[ALLOCA]], ptr addrspace(4) noundef align 16 dereferenceable(32) [[KERNARG_SEGMENT_PTR]], i64 32, i1 false)
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds [32 x i8], ptr addrspace(5) [[ALLOCA]], i32 0, i32 [[IDX:%.*]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(5) [[GEP]], align 1
; CHECK-NEXT:    store i8 [[LOAD]], ptr addrspace(1) [[OUT:%.*]], align 1
; CHECK-NEXT:    ret void
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  %kernarg.segment.ptr = call dereferenceable(32) align 16 ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr()
  call void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) %alloca, ptr addrspace(4) %kernarg.segment.ptr, i64 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  store i8 %load, ptr addrspace(1) %out
  ret void
}

; Alloca is written through a flat pointer
define i8 @memcpy_constant_arg_ptr_to_alloca_addrspacecast_to_flat(ptr addrspace(4) noalias readonly align 4 dereferenceable(32) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca_addrspacecast_to_flat(
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr [32 x i8], ptr addrspace(4) [[ARG:%.*]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr addrspace(4) [[GEP]], align 1
; CHECK-NEXT:    ret i8 [[LOAD]]
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  %alloca.cast.asc = addrspacecast ptr addrspace(5) %alloca to ptr
  call void @llvm.memcpy.p0.p4.i64(ptr %alloca.cast.asc, ptr addrspace(4) %arg, i64 31, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr addrspace(5) %alloca, i32 0, i32 %idx
  %load = load i8, ptr addrspace(5) %gep
  ret i8 %load
}

; Alloca is only addressed through flat pointer.
define i8 @memcpy_constant_arg_ptr_to_alloca_addrspacecast_to_flat2(ptr addrspace(4) noalias readonly align 4 dereferenceable(32) %arg, i32 %idx) {
; CHECK-LABEL: @memcpy_constant_arg_ptr_to_alloca_addrspacecast_to_flat2(
; CHECK-NEXT:    [[ALLOCA_CAST_ASC:%.*]] = addrspacecast ptr addrspace(4) [[ARG:%.*]] to ptr
; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[IDX:%.*]] to i64
; CHECK-NEXT:    [[GEP:%.*]] = getelementptr inbounds [32 x i8], ptr [[ALLOCA_CAST_ASC]], i64 0, i64 [[TMP1]]
; CHECK-NEXT:    [[LOAD:%.*]] = load i8, ptr [[GEP]], align 1
; CHECK-NEXT:    ret i8 [[LOAD]]
;
  %alloca = alloca [32 x i8], align 4, addrspace(5)
  %alloca.cast.asc = addrspacecast ptr addrspace(5) %alloca to ptr
  call void @llvm.memcpy.p0.p4.i64(ptr %alloca.cast.asc, ptr addrspace(4) %arg, i64 32, i1 false)
  %gep = getelementptr inbounds [32 x i8], ptr %alloca.cast.asc, i32 0, i32 %idx
  %load = load i8, ptr %gep
  ret i8 %load
}

%struct.ty = type { [4 x i32] }

define amdgpu_kernel void @byref_infloop(ptr %scratch, ptr addrspace(4) byref(%struct.ty) align 4 %arg) local_unnamed_addr #1 {
; CHECK-LABEL: @byref_infloop(
; CHECK-NEXT:  bb:
; CHECK-NEXT:    call void @llvm.memcpy.p0.p4.i32(ptr noundef nonnull align 4 dereferenceable(16) [[SCRATCH:%.*]], ptr addrspace(4) noundef align 4 dereferenceable(16) [[ARG:%.*]], i32 16, i1 false)
; CHECK-NEXT:    ret void
;
bb:
  %alloca = alloca [4 x i32], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i32(ptr addrspace(5) align 4 %alloca, ptr addrspace(4) align 4 %arg, i32 16, i1 false)
  call void @llvm.memcpy.p0.p5.i32(ptr align 4 %scratch, ptr addrspace(5) align 4 %alloca, i32 16, i1 false)
  ret void
}

define amdgpu_kernel void @byref_infloop_metadata(ptr %scratch, ptr addrspace(4) byref(%struct.ty) align 4 %arg) local_unnamed_addr #1 {
; CHECK-LABEL: @byref_infloop_metadata(
; CHECK-NEXT:  bb:
; CHECK-NEXT:    call void @llvm.memcpy.p0.p4.i32(ptr noundef nonnull align 4 dereferenceable(16) [[SCRATCH:%.*]], ptr addrspace(4) noundef align 4 dereferenceable(16) [[ARG:%.*]], i32 16, i1 false), !noalias !0
; CHECK-NEXT:    ret void
;
bb:
  %alloca = alloca [4 x i32], align 4, addrspace(5)
  call void @llvm.memcpy.p5.p4.i32(ptr addrspace(5) align 4 %alloca, ptr addrspace(4) align 4 %arg, i32 16, i1 false), !noalias !0
  call void @llvm.memcpy.p0.p5.i32(ptr align 4 %scratch, ptr addrspace(5) align 4 %alloca, i32 16, i1 false), !noalias !0
  ret void
}

define amdgpu_kernel void @byref_infloop_addrspacecast(ptr %scratch, ptr addrspace(4) byref(%struct.ty) align 4 %arg) local_unnamed_addr #1 {
; CHECK-LABEL: @byref_infloop_addrspacecast(
; CHECK-NEXT:  bb:
; CHECK-NEXT:    [[ADDRSPACECAST_ALLOCA:%.*]] = addrspacecast ptr addrspace(4) [[ARG:%.*]] to ptr
; CHECK-NEXT:    call void @llvm.memcpy.p0.p0.i64(ptr noundef nonnull align 4 dereferenceable(16) [[SCRATCH:%.*]], ptr noundef nonnull align 4 dereferenceable(16) [[ADDRSPACECAST_ALLOCA]], i64 16, i1 false)
; CHECK-NEXT:    ret void
;
bb:
  %alloca = alloca [4 x i32], align 4, addrspace(5)
  %addrspacecast.alloca = addrspacecast ptr addrspace(5) %alloca to ptr
  call void @llvm.memcpy.p0.p4.i64(ptr nonnull align 4 dereferenceable(16) %addrspacecast.alloca, ptr addrspace(4) align 4 dereferenceable(16) %arg, i64 16, i1 false)
  call void @llvm.memcpy.p0.p0.i64(ptr nonnull align 4 dereferenceable(16) %scratch, ptr nonnull align 4 dereferenceable(16) %addrspacecast.alloca, i64 16, i1 false)
  ret void
}

define amdgpu_kernel void @byref_infloop_memmove(ptr %scratch, ptr addrspace(4) byref(%struct.ty) align 4 %arg) local_unnamed_addr #1 {
; CHECK-LABEL: @byref_infloop_memmove(
; CHECK-NEXT:  bb:
; CHECK-NEXT:    call void @llvm.memmove.p0.p4.i32(ptr noundef nonnull align 4 dereferenceable(16) [[SCRATCH:%.*]], ptr addrspace(4) noundef align 4 dereferenceable(16) [[ARG:%.*]], i32 16, i1 false)
; CHECK-NEXT:    ret void
;
bb:
  %alloca = alloca [4 x i32], align 4, addrspace(5)
  call void @llvm.memmove.p5.p4.i32(ptr addrspace(5) align 4 %alloca, ptr addrspace(4) align 4 %arg, i32 16, i1 false)
  call void @llvm.memmove.p0.p5.i32(ptr align 4 %scratch, ptr addrspace(5) align 4 %alloca, i32 16, i1 false)
  ret void
}

declare void @llvm.memcpy.p0.p5.i32(ptr noalias nocapture writeonly, ptr addrspace(5) noalias nocapture readonly, i32, i1 immarg) #0
declare void @llvm.memcpy.p5.p4.i32(ptr addrspace(5) nocapture, ptr addrspace(4) nocapture, i32, i1) #0
declare void @llvm.memcpy.p0.p4.i64(ptr nocapture, ptr addrspace(4) nocapture, i64, i1) #0
declare void @llvm.memcpy.p0.p0.i64(ptr noalias nocapture writeonly, ptr noalias nocapture readonly, i64, i1 immarg) #0
declare void @llvm.memcpy.p5.p4.i64(ptr addrspace(5) nocapture, ptr addrspace(4) nocapture, i64, i1) #0
declare void @llvm.memmove.p5.p4.i32(ptr addrspace(5) nocapture, ptr addrspace(4) nocapture, i32, i1) #0
declare void @llvm.memmove.p0.p5.i32(ptr nocapture, ptr addrspace(5) nocapture, i32, i1) #0
declare ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr() #1

attributes #0 = { argmemonly nounwind willreturn }
attributes #1 = { nounwind readnone speculatable }

!0 = !{!1}
!1 = !{!1, !2}
!2 = !{!2}