atomicrmw_uinc_wrap.ll | Explore in Territory

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=bonaire -mattr=+flat-for-global -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,CI %s
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=tonga -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,VI %s
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,GFX9 %s
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1010 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GCN,GFX10 %s
; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1100 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefixes=GFX11 %s

; FIXME: Merge with other test. DS offset folding doesn't work due to
; register bank copies, and no return optimization is missing.

@lds0 = internal addrspace(3) global [512 x i32] undef, align 4
@lds1 = internal addrspace(3) global [512 x i64] undef, align 8

declare i32 @llvm.amdgcn.workitem.id.x() #0

define amdgpu_kernel void @lds_atomic_inc_ret_i32(ptr addrspace(1) %out, ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_ret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s2, s[6:7], 0x2
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v1, s2
; CI-NEXT:    ds_inc_rtn_u32 v2, v1, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_ret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s2, s[6:7], 0x8
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s2
; VI-NEXT:    ds_inc_rtn_u32 v2, v1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_ret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s2, s[6:7], 0x8
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_ret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x8
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_ret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x8
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 42 :: v_dual_mov_b32 v0, s0
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    global_store_b32 v1, v0, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_ret_i32_offset(ptr addrspace(1) %out, ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_ret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s2, s[6:7], 0x2
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v1, s2
; CI-NEXT:    ds_inc_rtn_u32 v2, v1, v0 offset:16
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_ret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s2, s[6:7], 0x8
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s2
; VI-NEXT:    ds_inc_rtn_u32 v2, v1, v0 offset:16
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_ret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s2, s[6:7], 0x8
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s2
; GFX9-NEXT:    ds_inc_rtn_u32 v0, v1, v0 offset:16
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_ret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x8
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v1, s0
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u32 v0, v1, v0 offset:16
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_ret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x8
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, s0
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u32 v0, v1, v0 offset:16
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    global_store_b32 v1, v0, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(3) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(3) %gep, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_noret_i32(ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_noret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s0, s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v1, s0
; CI-NEXT:    ds_inc_u32 v1, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_noret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s0, s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s0
; VI-NEXT:    ds_inc_u32 v1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_noret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    ds_inc_u32 v0, v1
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_noret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    ds_inc_u32 v0, v1
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_noret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 42 :: v_dual_mov_b32 v0, s0
; GFX11-NEXT:    ds_inc_u32 v0, v1
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_noret_i32_offset(ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_noret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s0, s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v1, s0
; CI-NEXT:    ds_inc_u32 v1, v0 offset:16
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_noret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s0, s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s0
; VI-NEXT:    ds_inc_u32 v1, v0 offset:16
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_noret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s0
; GFX9-NEXT:    ds_inc_u32 v1, v0 offset:16
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_noret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v1, s0
; GFX10-NEXT:    ds_inc_u32 v1, v0 offset:16
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_noret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, s0
; GFX11-NEXT:    ds_inc_u32 v1, v0 offset:16
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(3) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(3) %gep, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i32(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v0, v1, v0, s[2:3] glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b32 v1, v0, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(1) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i32_offset(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 16
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 16
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] offset:16 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] offset:16 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v0, v1, v0, s[2:3] offset:16 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b32 v1, v0, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i32_offset_sistem(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i32_offset_sistem:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 16
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i32_offset_sistem:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 16
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i32_offset_sistem:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] offset:16 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i32_offset_sistem:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v0, v1, v0, s[2:3] offset:16 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dword v1, v0, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i32_offset_sistem:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v0, v1, v0, s[2:3] offset:16 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b32 v1, v0, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i32(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v1, v0, s[0:1]
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v1, v0, s[0:1]
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v1, v0, s[0:1]
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(1) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i32_offset(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 16
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 16
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v1, v0, s[0:1] offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v1, v0, s[0:1] offset:16
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v1, v0, s[0:1] offset:16
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i32_offset_system(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i32_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 16
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i32_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 16
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i32_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v1, v0, s[0:1] offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i32_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v1, v0, s[0:1] offset:16
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i32_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v0, 42 :: v_dual_mov_b32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v1, v0, s[0:1] offset:16
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 seq_cst, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i32_offset_addr64(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i32_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; CI-NEXT:    v_mov_b32_e32 v3, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_add_i32_e32 v0, vcc, 20, v0
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    flat_atomic_inc v3, v[0:1], v3 glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    flat_store_dword v[0:1], v3
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i32_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; VI-NEXT:    v_mov_b32_e32 v3, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v0, vcc, 20, v0
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    flat_atomic_inc v3, v[0:1], v3 glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    flat_store_dword v[0:1], v3
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i32_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v1, v0, v1, s[2:3] offset:20 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dword v0, v1, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i32_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v1, v0, v1, s[2:3] offset:20 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dword v0, v1, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i32_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v1, 42 :: v_dual_and_b32 v0, 0x3ff, v0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1)
; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v1, v0, v1, s[2:3] offset:20 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b32 v0, v1, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i32, ptr addrspace(1) %ptr, i32 %id
  %out.gep = getelementptr i32, ptr addrspace(1) %out, i32 %id
  %gep = getelementptr i32, ptr addrspace(1) %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr addrspace(1) %out.gep, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i32_offset_addr64(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i32_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_add_i32_e32 v0, vcc, 20, v0
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i32_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v0, vcc, 20, v0
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i32_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc v0, v1, s[0:1] offset:20
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i32_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc v0, v1, s[0:1] offset:20
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i32_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v1, 42 :: v_dual_and_b32 v0, 0x3ff, v0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1)
; GFX11-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u32 v0, v1, s[0:1] offset:20
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i32, ptr addrspace(1) %ptr, i32 %id
  %gep = getelementptr i32, ptr addrspace(1) %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @atomic_inc_shl_base_lds_0_i32(ptr addrspace(1) %out, ptr addrspace(1) %add_use) #1 {
; CI-LABEL: atomic_inc_shl_base_lds_0_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v1, 2, v0
; CI-NEXT:    v_mov_b32_e32 v2, 9
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    ds_inc_rtn_u32 v2, v1, v2 offset:8
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_add_i32_e32 v3, vcc, 2, v0
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_store_dword v[0:1], v3
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: atomic_inc_shl_base_lds_0_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v1, 2, v0
; VI-NEXT:    v_mov_b32_e32 v2, 9
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    ds_inc_rtn_u32 v2, v1, v2 offset:8
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_add_u32_e32 v3, vcc, 2, v0
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_store_dword v[0:1], v3
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: atomic_inc_shl_base_lds_0_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v1, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v2, 9
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    ds_inc_rtn_u32 v1, v1, v2 offset:8
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_add_u32_e32 v0, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    global_store_dword v2, v0, s[2:3]
; GFX9-NEXT:    global_store_dword v2, v1, s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: atomic_inc_shl_base_lds_0_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    v_lshlrev_b32_e32 v1, 2, v0
; GFX10-NEXT:    v_mov_b32_e32 v2, 9
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_add_nc_u32_e32 v0, 2, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u32 v1, v1, v2 offset:8
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    global_store_dword v2, v0, s[2:3]
; GFX10-NEXT:    global_store_dword v2, v1, s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: atomic_inc_shl_base_lds_0_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1)
; GFX11-NEXT:    v_dual_mov_b32 v2, 9 :: v_dual_lshlrev_b32 v1, 2, v0
; GFX11-NEXT:    v_add_nc_u32_e32 v0, 2, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u32 v1, v1, v2 offset:8
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v2, 0
; GFX11-NEXT:    s_clause 0x1
; GFX11-NEXT:    global_store_b32 v2, v0, s[2:3]
; GFX11-NEXT:    global_store_b32 v2, v1, s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %tid.x = tail call i32 @llvm.amdgcn.workitem.id.x() #2
  %idx.0 = add nsw i32 %tid.x, 2
  %arrayidx0 = getelementptr inbounds [512 x i32], ptr addrspace(3) @lds0, i32 0, i32 %idx.0
  %result = atomicrmw uinc_wrap ptr addrspace(3) %arrayidx0, i32 9 syncscope("agent") seq_cst, align 4
  store i32 %idx.0, ptr addrspace(1) %add_use, align 4
  store i32 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_ret_i64(ptr addrspace(1) %out, ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_ret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s2, s[6:7], 0x2
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1]
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_ret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s2, s[6:7], 0x8
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1]
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_ret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s2, s[6:7], 0x8
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s2
; GFX9-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1]
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_ret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x8
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_ret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x8
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1]
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v2, 0
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_ret_i64_offset(ptr addrspace(1) %out, ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_ret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s2, s[6:7], 0x2
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1] offset:32
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_ret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s2, s[6:7], 0x8
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1] offset:32
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_ret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s2, s[6:7], 0x8
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s2
; GFX9-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1] offset:32
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_ret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x8
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1] offset:32
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_ret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x8
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u64 v[0:1], v2, v[0:1] offset:32
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v2, 0
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(3) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(3) %gep, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_noret_i64(ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_noret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s0, s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    ds_inc_u64 v2, v[0:1]
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_noret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s0, s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    ds_inc_u64 v2, v[0:1]
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_noret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    ds_inc_u64 v2, v[0:1]
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_noret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    ds_inc_u64 v2, v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_noret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    ds_inc_u64 v2, v[0:1]
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @lds_atomic_inc_noret_i64_offset(ptr addrspace(3) %ptr) #1 {
; CI-LABEL: lds_atomic_inc_noret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s0, s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    ds_inc_u64 v2, v[0:1] offset:32
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_endpgm
;
; VI-LABEL: lds_atomic_inc_noret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s0, s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    ds_inc_u64 v2, v[0:1] offset:32
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: lds_atomic_inc_noret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    ds_inc_u64 v2, v[0:1] offset:32
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: lds_atomic_inc_noret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    ds_inc_u64 v2, v[0:1] offset:32
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: lds_atomic_inc_noret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    ds_inc_u64 v2, v[0:1] offset:32
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(3) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(3) %gep, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i64(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v[0:1], v2, v[0:1], s[2:3] glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(1) %ptr, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i64_offset(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 32
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 32
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i64_offset_system(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i64_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 32
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i64_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 32
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i64_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i64_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dwordx2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i64_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v[0:1], v2, v[0:1], s[2:3] offset:32 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i64(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1]
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1]
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr addrspace(1) %ptr, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i64_offset(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 32
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 32
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1] offset:32
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1] offset:32
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v2, v[0:1], s[0:1] offset:32
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i64_offset_system(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i64_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 32
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i64_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 32
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i64_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1] offset:32
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i64_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v2, v[0:1], s[0:1] offset:32
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i64_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v2, v[0:1], s[0:1] offset:32
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr addrspace(1) %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 seq_cst, align 8
  ret void
}

define amdgpu_kernel void @global_atomic_inc_ret_i64_offset_addr64(ptr addrspace(1) %out, ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_ret_i64_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    v_add_i32_e32 v2, vcc, v0, v4
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_add_i32_e32 v2, vcc, 40, v2
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    v_add_i32_e32 v2, vcc, v2, v4
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_ret_i64_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    v_add_u32_e32 v2, vcc, v0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_add_u32_e32 v2, vcc, 40, v2
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    v_add_u32_e32 v2, vcc, v2, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_ret_i64_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v[0:1], v3, v[1:2], s[2:3] offset:40 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    global_store_dwordx2 v3, v[0:1], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_ret_i64_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v[0:1], v3, v[1:2], s[2:3] offset:40 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    global_store_dwordx2 v3, v[0:1], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_ret_i64_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_and_b32_e32 v2, 0x3ff, v0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_lshlrev_b32 v2, 3, v2
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v[0:1], v2, v[0:1], s[2:3] offset:40 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    global_store_b64 v2, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i64, ptr addrspace(1) %ptr, i32 %id
  %out.gep = getelementptr i64, ptr addrspace(1) %out, i32 %id
  %gep = getelementptr i64, ptr addrspace(1) %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr addrspace(1) %out.gep, align 4
  ret void
}

define amdgpu_kernel void @global_atomic_inc_noret_i64_offset_addr64(ptr addrspace(1) %ptr) #1 {
; CI-LABEL: global_atomic_inc_noret_i64_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 3, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v2, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_add_i32_e32 v2, vcc, 40, v2
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: global_atomic_inc_noret_i64_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_add_u32_e32 v2, vcc, 40, v2
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: global_atomic_inc_noret_i64_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    v_lshlrev_b32_e32 v0, 3, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    global_atomic_inc_x2 v0, v[1:2], s[0:1] offset:40
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: global_atomic_inc_noret_i64_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    v_lshlrev_b32_e32 v0, 3, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    global_atomic_inc_x2 v0, v[1:2], s[0:1] offset:40
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: global_atomic_inc_noret_i64_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_and_b32_e32 v2, 0x3ff, v0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_lshlrev_b32 v2, 3, v2
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    global_atomic_inc_u64 v2, v[0:1], s[0:1] offset:40
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i64, ptr addrspace(1) %ptr, i32 %id
  %gep = getelementptr i64, ptr addrspace(1) %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr addrspace(1) %gep, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i32(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    v_mov_b32_e32 v1, s3
; GFX9-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_store_dword v[0:1], v2
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s2
; GFX10-NEXT:    v_mov_b32_e32 v1, s3
; GFX10-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_store_dword v[0:1], v2
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX11-NEXT:    flat_atomic_inc_u32 v2, v[0:1], v2 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_store_b32 v[0:1], v2
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr %ptr, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i32_offset(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 16
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 16
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    v_mov_b32_e32 v1, s3
; GFX9-NEXT:    flat_atomic_inc v2, v[0:1], v2 offset:16 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_store_dword v[0:1], v2
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s2, s2, 16
; GFX10-NEXT:    s_addc_u32 s3, s3, 0
; GFX10-NEXT:    v_mov_b32_e32 v0, s2
; GFX10-NEXT:    v_mov_b32_e32 v1, s3
; GFX10-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_store_dword v[0:1], v2
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX11-NEXT:    flat_atomic_inc_u32 v2, v[0:1], v2 offset:16 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_store_b32 v[0:1], v2
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i32_offset_system(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i32_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 16
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i32_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 16
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i32_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    v_mov_b32_e32 v1, s3
; GFX9-NEXT:    flat_atomic_inc v2, v[0:1], v2 offset:16 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_store_dword v[0:1], v2
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i32_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s2, s2, 16
; GFX10-NEXT:    s_addc_u32 s3, s3, 0
; GFX10-NEXT:    v_mov_b32_e32 v0, s2
; GFX10-NEXT:    v_mov_b32_e32 v1, s3
; GFX10-NEXT:    flat_atomic_inc v2, v[0:1], v2 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_store_dword v[0:1], v2
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i32_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX11-NEXT:    flat_atomic_inc_u32 v2, v[0:1], v2 offset:16 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_store_b32 v[0:1], v2
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 seq_cst, align 4
  store i32 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i32(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_atomic_inc v[0:1], v2
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_atomic_inc v[0:1], v2
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_atomic_inc_u32 v[0:1], v2
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr %ptr, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i32_offset(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i32_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 16
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i32_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 16
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i32_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_atomic_inc v[0:1], v2 offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i32_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s0, s0, 16
; GFX10-NEXT:    s_addc_u32 s1, s1, 0
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_atomic_inc v[0:1], v2
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i32_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_atomic_inc_u32 v[0:1], v2 offset:16
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i32_offset_system(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i32_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 16
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i32_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 16
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i32_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    flat_atomic_inc v[0:1], v2 offset:16
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i32_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s0, s0, 16
; GFX10-NEXT:    s_addc_u32 s1, s1, 0
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    flat_atomic_inc v[0:1], v2
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i32_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    flat_atomic_inc_u32 v[0:1], v2 offset:16
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i32, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 seq_cst, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i32_offset_addr64(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i32_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; CI-NEXT:    v_mov_b32_e32 v3, 42
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_add_i32_e32 v0, vcc, 20, v0
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    flat_atomic_inc v3, v[0:1], v3 glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    flat_store_dword v[0:1], v3
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i32_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; VI-NEXT:    v_mov_b32_e32 v3, 42
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v0, vcc, 20, v0
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    flat_atomic_inc v3, v[0:1], v3 glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    flat_store_dword v[0:1], v3
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i32_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v3, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s2
; GFX9-NEXT:    v_mov_b32_e32 v1, s3
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, v0, v2
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    flat_atomic_inc v3, v[0:1], v3 offset:20 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, v0, v2
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    flat_store_dword v[0:1], v3
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i32_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX10-NEXT:    v_mov_b32_e32 v3, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s2
; GFX10-NEXT:    v_mov_b32_e32 v1, s3
; GFX10-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX10-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 20
; GFX10-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    flat_atomic_inc v3, v[0:1], v3 glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX10-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    flat_store_dword v[0:1], v3
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i32_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_dual_mov_b32 v3, 42 :: v_dual_and_b32 v0, 0x3ff, v0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1)
; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX11-NEXT:    flat_atomic_inc_u32 v3, v[0:1], v3 offset:20 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX11-NEXT:    flat_store_b32 v[0:1], v3
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i32, ptr %ptr, i32 %id
  %out.gep = getelementptr i32, ptr %out, i32 %id
  %gep = getelementptr i32, ptr %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result, ptr %out.gep, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i32_offset_addr64(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i32_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v0, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_add_i32_e32 v0, vcc, 20, v0
; CI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v2, 42
; CI-NEXT:    flat_atomic_inc v[0:1], v2
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i32_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v0, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v0, vcc, 20, v0
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v2, 42
; VI-NEXT:    flat_atomic_inc v[0:1], v2
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i32_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s0
; GFX9-NEXT:    v_mov_b32_e32 v1, s1
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, v0, v2
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v2, 42
; GFX9-NEXT:    flat_atomic_inc v[0:1], v2 offset:20
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i32_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX10-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    v_mov_b32_e32 v2, 42
; GFX10-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 20
; GFX10-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    flat_atomic_inc v[0:1], v2
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i32_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1)
; GFX11-NEXT:    v_lshlrev_b32_e32 v2, 2, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v2
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    flat_atomic_inc_u32 v[0:1], v2 offset:20
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i32, ptr %ptr, i32 %id
  %gep = getelementptr i32, ptr %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr %gep, i32 42 syncscope("agent") seq_cst, align 4
  ret void
}

define amdgpu_kernel void @atomic_inc_shl_base_lds_0_i64(ptr addrspace(1) %out, ptr addrspace(1) %add_use) #1 {
; CI-LABEL: atomic_inc_shl_base_lds_0_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v1, 9
; CI-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; CI-NEXT:    v_mov_b32_e32 v2, 0
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    ds_inc_rtn_u64 v[1:2], v3, v[1:2] offset:16
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v4, s3
; CI-NEXT:    v_add_i32_e32 v0, vcc, 2, v0
; CI-NEXT:    v_mov_b32_e32 v3, s2
; CI-NEXT:    flat_store_dword v[3:4], v0
; CI-NEXT:    v_mov_b32_e32 v4, s1
; CI-NEXT:    v_mov_b32_e32 v3, s0
; CI-NEXT:    flat_store_dwordx2 v[3:4], v[1:2]
; CI-NEXT:    s_endpgm
;
; VI-LABEL: atomic_inc_shl_base_lds_0_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v1, 9
; VI-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; VI-NEXT:    v_mov_b32_e32 v2, 0
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    ds_inc_rtn_u64 v[1:2], v3, v[1:2] offset:16
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v4, s3
; VI-NEXT:    v_add_u32_e32 v0, vcc, 2, v0
; VI-NEXT:    v_mov_b32_e32 v3, s2
; VI-NEXT:    flat_store_dword v[3:4], v0
; VI-NEXT:    v_mov_b32_e32 v4, s1
; VI-NEXT:    v_mov_b32_e32 v3, s0
; VI-NEXT:    flat_store_dwordx2 v[3:4], v[1:2]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: atomic_inc_shl_base_lds_0_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    v_mov_b32_e32 v1, 9
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    ds_inc_rtn_u64 v[1:2], v3, v[1:2] offset:16
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_add_u32_e32 v0, 2, v0
; GFX9-NEXT:    v_mov_b32_e32 v3, 0
; GFX9-NEXT:    global_store_dword v3, v0, s[2:3]
; GFX9-NEXT:    global_store_dwordx2 v3, v[1:2], s[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: atomic_inc_shl_base_lds_0_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    v_mov_b32_e32 v1, 9
; GFX10-NEXT:    v_mov_b32_e32 v2, 0
; GFX10-NEXT:    v_lshlrev_b32_e32 v3, 3, v0
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_add_nc_u32_e32 v0, 2, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u64 v[1:2], v3, v[1:2] offset:16
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v3, 0
; GFX10-NEXT:    global_store_dword v3, v0, s[2:3]
; GFX10-NEXT:    global_store_dwordx2 v3, v[1:2], s[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: atomic_inc_shl_base_lds_0_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    v_and_b32_e32 v2, 0x3ff, v0
; GFX11-NEXT:    v_mov_b32_e32 v0, 9
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_3)
; GFX11-NEXT:    v_lshlrev_b32_e32 v3, 3, v2
; GFX11-NEXT:    v_add_nc_u32_e32 v2, 2, v2
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u64 v[0:1], v3, v[0:1] offset:16
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v3, 0
; GFX11-NEXT:    s_clause 0x1
; GFX11-NEXT:    global_store_b32 v3, v2, s[2:3]
; GFX11-NEXT:    global_store_b64 v3, v[0:1], s[0:1]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %tid.x = tail call i32 @llvm.amdgcn.workitem.id.x() #2
  %idx.0 = add nsw i32 %tid.x, 2
  %arrayidx0 = getelementptr inbounds [512 x i64], ptr addrspace(3) @lds1, i32 0, i32 %idx.0
  %result = atomicrmw uinc_wrap ptr addrspace(3) %arrayidx0, i64 9 syncscope("agent") seq_cst, align 8
  store i32 %idx.0, ptr addrspace(1) %add_use, align 4
  store i64 %result, ptr addrspace(1) %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i64(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    s_add_u32 s0, s0, 4
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v5, s1
; CI-NEXT:    v_mov_b32_e32 v4, s0
; CI-NEXT:    flat_store_dword v[2:3], v0
; CI-NEXT:    flat_store_dword v[4:5], v1
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    s_add_u32 s0, s0, 4
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v5, s1
; VI-NEXT:    v_mov_b32_e32 v4, s0
; VI-NEXT:    flat_store_dword v[2:3], v0
; VI-NEXT:    flat_store_dword v[4:5], v1
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s2
; GFX9-NEXT:    v_mov_b32_e32 v3, s3
; GFX9-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v2, s2
; GFX10-NEXT:    v_mov_b32_e32 v3, s3
; GFX10-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s2
; GFX11-NEXT:    v_mov_b32_e32 v3, s3
; GFX11-NEXT:    flat_atomic_inc_u64 v[0:1], v[2:3], v[0:1] glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_store_b64 v[2:3], v[0:1]
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr %ptr, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i64_offset(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 32
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    s_add_u32 s0, s0, 4
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v5, s1
; CI-NEXT:    v_mov_b32_e32 v4, s0
; CI-NEXT:    flat_store_dword v[2:3], v0
; CI-NEXT:    flat_store_dword v[4:5], v1
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 32
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    s_add_u32 s0, s0, 4
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v5, s1
; VI-NEXT:    v_mov_b32_e32 v4, s0
; VI-NEXT:    flat_store_dword v[2:3], v0
; VI-NEXT:    flat_store_dword v[4:5], v1
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s2
; GFX9-NEXT:    v_mov_b32_e32 v3, s3
; GFX9-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] offset:32 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s2, s2, 32
; GFX10-NEXT:    s_addc_u32 s3, s3, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, s2
; GFX10-NEXT:    v_mov_b32_e32 v3, s3
; GFX10-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s2
; GFX11-NEXT:    v_mov_b32_e32 v3, s3
; GFX11-NEXT:    flat_atomic_inc_u64 v[0:1], v[2:3], v[0:1] offset:32 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_store_b64 v[2:3], v[0:1]
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i64_offset_system(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i64_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s2, s2, 32
; CI-NEXT:    s_addc_u32 s3, s3, 0
; CI-NEXT:    v_mov_b32_e32 v2, s2
; CI-NEXT:    v_mov_b32_e32 v3, s3
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    s_add_u32 s0, s0, 4
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v5, s1
; CI-NEXT:    v_mov_b32_e32 v4, s0
; CI-NEXT:    flat_store_dword v[2:3], v0
; CI-NEXT:    flat_store_dword v[4:5], v1
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i64_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s2, s2, 32
; VI-NEXT:    s_addc_u32 s3, s3, 0
; VI-NEXT:    v_mov_b32_e32 v2, s2
; VI-NEXT:    v_mov_b32_e32 v3, s3
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    s_add_u32 s0, s0, 4
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v5, s1
; VI-NEXT:    v_mov_b32_e32 v4, s0
; VI-NEXT:    flat_store_dword v[2:3], v0
; VI-NEXT:    flat_store_dword v[4:5], v1
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i64_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v2, s2
; GFX9-NEXT:    v_mov_b32_e32 v3, s3
; GFX9-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] offset:32 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i64_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s2, s2, 32
; GFX10-NEXT:    s_addc_u32 s3, s3, 0
; GFX10-NEXT:    v_mov_b32_e32 v2, s2
; GFX10-NEXT:    v_mov_b32_e32 v3, s3
; GFX10-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i64_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s2
; GFX11-NEXT:    v_mov_b32_e32 v3, s3
; GFX11-NEXT:    flat_atomic_inc_u64 v[0:1], v[2:3], v[0:1] offset:32 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_store_b64 v[2:3], v[0:1]
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 seq_cst, align 8
  store i64 %result, ptr %out, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i64(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_atomic_inc_u64 v[2:3], v[0:1]
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %result = atomicrmw uinc_wrap ptr %ptr, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i64_offset(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i64_offset:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 32
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i64_offset:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 32
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i64_offset:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1] offset:32
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i64_offset:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s0, s0, 32
; GFX10-NEXT:    s_addc_u32 s1, s1, 0
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i64_offset:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_atomic_inc_u64 v[2:3], v[0:1] offset:32
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i64_offset_system(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i64_offset_system:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    s_add_u32 s0, s0, 32
; CI-NEXT:    s_addc_u32 s1, s1, 0
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i64_offset_system:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    s_add_u32 s0, s0, 32
; VI-NEXT:    s_addc_u32 s1, s1, 0
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i64_offset_system:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v0, 42
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1] offset:32
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i64_offset_system:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_add_u32 s0, s0, 32
; GFX10-NEXT:    s_addc_u32 s1, s1, 0
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i64_offset_system:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_mov_b32_e32 v0, 42
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    flat_atomic_inc_u64 v[2:3], v[0:1] offset:32
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %gep = getelementptr i64, ptr %ptr, i32 4
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 seq_cst, align 8
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_ret_i64_offset_addr64(ptr %out, ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_ret_i64_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    v_add_i32_e32 v2, vcc, v0, v4
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_add_i32_e32 v2, vcc, 40, v2
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    v_mov_b32_e32 v3, s1
; CI-NEXT:    v_mov_b32_e32 v2, s0
; CI-NEXT:    v_add_i32_e32 v2, vcc, v2, v4
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    v_add_i32_e32 v4, vcc, 4, v2
; CI-NEXT:    v_addc_u32_e32 v5, vcc, 0, v3, vcc
; CI-NEXT:    flat_store_dword v[2:3], v0
; CI-NEXT:    flat_store_dword v[4:5], v1
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_ret_i64_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    v_add_u32_e32 v2, vcc, v0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_add_u32_e32 v2, vcc, 40, v2
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_mov_b32_e32 v2, s0
; VI-NEXT:    v_add_u32_e32 v2, vcc, v2, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    v_add_u32_e32 v4, vcc, 4, v2
; VI-NEXT:    v_addc_u32_e32 v5, vcc, 0, v3, vcc
; VI-NEXT:    flat_store_dword v[2:3], v0
; VI-NEXT:    flat_store_dword v[4:5], v1
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_ret_i64_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v5, 3, v0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v4, s3
; GFX9-NEXT:    v_mov_b32_e32 v3, s2
; GFX9-NEXT:    v_add_co_u32_e32 v3, vcc, v3, v5
; GFX9-NEXT:    v_addc_co_u32_e32 v4, vcc, 0, v4, vcc
; GFX9-NEXT:    flat_atomic_inc_x2 v[0:1], v[3:4], v[1:2] offset:40 glc
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, v2, v5
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_ret_i64_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s2
; GFX10-NEXT:    v_mov_b32_e32 v1, s3
; GFX10-NEXT:    v_add_co_u32 v2, vcc_lo, v0, v4
; GFX10-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 40
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
; GFX10-NEXT:    flat_atomic_inc_x2 v[0:1], v[2:3], v[0:1] glc
; GFX10-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v3, s1
; GFX10-NEXT:    v_mov_b32_e32 v2, s0
; GFX10-NEXT:    v_add_co_u32 v2, vcc_lo, v2, v4
; GFX10-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
; GFX10-NEXT:    flat_store_dwordx2 v[2:3], v[0:1]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_ret_i64_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_1)
; GFX11-NEXT:    v_dual_mov_b32 v3, 0 :: v_dual_lshlrev_b32 v4, 3, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v4
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX11-NEXT:    flat_atomic_inc_u64 v[0:1], v[0:1], v[2:3] offset:40 glc
; GFX11-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_dual_mov_b32 v3, s1 :: v_dual_mov_b32 v2, s0
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_u32 v2, vcc_lo, v2, v4
; GFX11-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
; GFX11-NEXT:    flat_store_b64 v[2:3], v[0:1]
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i64, ptr %ptr, i32 %id
  %out.gep = getelementptr i64, ptr %out, i32 %id
  %gep = getelementptr i64, ptr %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 syncscope("agent") seq_cst, align 8
  store i64 %result, ptr %out.gep, align 4
  ret void
}

define amdgpu_kernel void @flat_atomic_inc_noret_i64_offset_addr64(ptr %ptr) #1 {
; CI-LABEL: flat_atomic_inc_noret_i64_offset_addr64:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; CI-NEXT:    v_lshlrev_b32_e32 v2, 3, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    v_add_i32_e32 v2, vcc, v0, v2
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    v_add_i32_e32 v2, vcc, 40, v2
; CI-NEXT:    v_mov_b32_e32 v1, 0
; CI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; CI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; CI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; CI-NEXT:    buffer_wbinvl1_vol
; CI-NEXT:    s_endpgm
;
; VI-LABEL: flat_atomic_inc_noret_i64_offset_addr64:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; VI-NEXT:    v_lshlrev_b32_e32 v2, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, v0, v2
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    v_add_u32_e32 v2, vcc, 40, v2
; VI-NEXT:    v_mov_b32_e32 v1, 0
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    buffer_wbinvl1_vol
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: flat_atomic_inc_noret_i64_offset_addr64:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX9-NEXT:    v_lshlrev_b32_e32 v0, 3, v0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v4, s1
; GFX9-NEXT:    v_mov_b32_e32 v3, s0
; GFX9-NEXT:    v_add_co_u32_e32 v3, vcc, v3, v0
; GFX9-NEXT:    v_addc_co_u32_e32 v4, vcc, 0, v4, vcc
; GFX9-NEXT:    flat_atomic_inc_x2 v[3:4], v[1:2] offset:40
; GFX9-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; GFX9-NEXT:    buffer_wbinvl1_vol
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: flat_atomic_inc_noret_i64_offset_addr64:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dwordx2 s[0:1], s[6:7], 0x0
; GFX10-NEXT:    v_lshlrev_b32_e32 v2, 3, v0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    v_mov_b32_e32 v1, s1
; GFX10-NEXT:    v_add_co_u32 v2, vcc_lo, v0, v2
; GFX10-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v1, vcc_lo
; GFX10-NEXT:    v_mov_b32_e32 v0, 42
; GFX10-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 40
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo
; GFX10-NEXT:    flat_atomic_inc_x2 v[2:3], v[0:1]
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX10-NEXT:    buffer_gl1_inv
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: flat_atomic_inc_noret_i64_offset_addr64:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b64 s[0:1], s[2:3], 0x0
; GFX11-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX11-NEXT:    v_mov_b32_e32 v2, 42
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_1)
; GFX11-NEXT:    v_dual_mov_b32 v3, 0 :: v_dual_lshlrev_b32 v4, 3, v0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX11-NEXT:    v_add_co_u32 v0, vcc_lo, v0, v4
; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX11-NEXT:    v_add_co_ci_u32_e32 v1, vcc_lo, 0, v1, vcc_lo
; GFX11-NEXT:    flat_atomic_inc_u64 v[0:1], v[2:3] offset:40
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    s_waitcnt_vscnt null, 0x0
; GFX11-NEXT:    buffer_gl1_inv
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    s_endpgm
  %id = call i32 @llvm.amdgcn.workitem.id.x()
  %gep.tid = getelementptr i64, ptr %ptr, i32 %id
  %gep = getelementptr i64, ptr %gep.tid, i32 5
  %result = atomicrmw uinc_wrap ptr %gep, i64 42 syncscope("agent") seq_cst, align 8
  ret void
}

define amdgpu_kernel void @nocse_lds_atomic_inc_ret_i32(ptr addrspace(1) %out0, ptr addrspace(1) %out1, ptr addrspace(3) %ptr) #1 {
; CI-LABEL: nocse_lds_atomic_inc_ret_i32:
; CI:       ; %bb.0:
; CI-NEXT:    s_load_dword s4, s[6:7], 0x4
; CI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; CI-NEXT:    v_mov_b32_e32 v0, 42
; CI-NEXT:    s_mov_b32 m0, -1
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v1, s4
; CI-NEXT:    ds_inc_rtn_u32 v2, v1, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    ds_inc_rtn_u32 v3, v1, v0
; CI-NEXT:    s_waitcnt lgkmcnt(0)
; CI-NEXT:    v_mov_b32_e32 v0, s0
; CI-NEXT:    v_mov_b32_e32 v1, s1
; CI-NEXT:    flat_store_dword v[0:1], v2
; CI-NEXT:    v_mov_b32_e32 v0, s2
; CI-NEXT:    v_mov_b32_e32 v1, s3
; CI-NEXT:    flat_store_dword v[0:1], v3
; CI-NEXT:    s_endpgm
;
; VI-LABEL: nocse_lds_atomic_inc_ret_i32:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dword s4, s[6:7], 0x10
; VI-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; VI-NEXT:    v_mov_b32_e32 v0, 42
; VI-NEXT:    s_mov_b32 m0, -1
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s4
; VI-NEXT:    ds_inc_rtn_u32 v2, v1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    ds_inc_rtn_u32 v3, v1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v0, s0
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    v_mov_b32_e32 v0, s2
; VI-NEXT:    v_mov_b32_e32 v1, s3
; VI-NEXT:    flat_store_dword v[0:1], v3
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: nocse_lds_atomic_inc_ret_i32:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dword s4, s[6:7], 0x10
; GFX9-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX9-NEXT:    v_mov_b32_e32 v1, 42
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v0, s4
; GFX9-NEXT:    ds_inc_rtn_u32 v2, v0, v1
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    global_store_dword v1, v2, s[0:1]
; GFX9-NEXT:    global_store_dword v1, v0, s[2:3]
; GFX9-NEXT:    s_endpgm
;
; GFX10-LABEL: nocse_lds_atomic_inc_ret_i32:
; GFX10:       ; %bb.0:
; GFX10-NEXT:    s_load_dword s0, s[6:7], 0x10
; GFX10-NEXT:    v_mov_b32_e32 v1, 42
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    v_mov_b32_e32 v0, s0
; GFX10-NEXT:    s_load_dwordx4 s[0:3], s[6:7], 0x0
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    ds_inc_rtn_u32 v2, v0, v1
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX10-NEXT:    s_waitcnt lgkmcnt(0)
; GFX10-NEXT:    buffer_gl0_inv
; GFX10-NEXT:    v_mov_b32_e32 v1, 0
; GFX10-NEXT:    global_store_dword v1, v2, s[0:1]
; GFX10-NEXT:    global_store_dword v1, v0, s[2:3]
; GFX10-NEXT:    s_endpgm
;
; GFX11-LABEL: nocse_lds_atomic_inc_ret_i32:
; GFX11:       ; %bb.0:
; GFX11-NEXT:    s_load_b32 s0, s[2:3], 0x10
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    v_dual_mov_b32 v1, 42 :: v_dual_mov_b32 v0, s0
; GFX11-NEXT:    s_load_b128 s[0:3], s[2:3], 0x0
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    ds_inc_rtn_u32 v2, v0, v1
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    ds_inc_rtn_u32 v0, v0, v1
; GFX11-NEXT:    s_waitcnt lgkmcnt(0)
; GFX11-NEXT:    buffer_gl0_inv
; GFX11-NEXT:    v_mov_b32_e32 v1, 0
; GFX11-NEXT:    s_clause 0x1
; GFX11-NEXT:    global_store_b32 v1, v2, s[0:1]
; GFX11-NEXT:    global_store_b32 v1, v0, s[2:3]
; GFX11-NEXT:    s_nop 0
; GFX11-NEXT:    s_sendmsg sendmsg(MSG_DEALLOC_VGPRS)
; GFX11-NEXT:    s_endpgm
  %result0 = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  %result1 = atomicrmw uinc_wrap ptr addrspace(3) %ptr, i32 42 syncscope("agent") seq_cst, align 4
  store i32 %result0, ptr addrspace(1) %out0, align 4
  store i32 %result1, ptr addrspace(1) %out1, align 4
  ret void
}

attributes #0 = { nounwind speculatable willreturn memory(none) }
attributes #1 = { nounwind }
attributes #2 = { nounwind memory(none) }
;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
; GCN: {{.*}}
llvm/llvm/test/CodeGen/AMDGPU/GlobalISel/atomicrmw_uinc_wrap.ll
