llvm/test/CodeGen/AMDGPU/memory-legalizer-flat-lastuse.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 < %s | FileCheck --check-prefixes=GFX12,GFX12-WGP %s
; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1200 -mattr=+cumode < %s | FileCheck --check-prefixes=GFX12,GFX12-CU %s

define amdgpu_kernel void @flat_last_use_load_0(ptr %in, ptr %out) {
; GFX12-LABEL: flat_last_use_load_0:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b128 s[0:3], s[0:1], 0x0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX12-NEXT:    flat_load_b32 v2, v[0:1] th:TH_LOAD_LU
; GFX12-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX12-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-NEXT:    flat_store_b32 v[0:1], v2
; GFX12-NEXT:    s_endpgm
entry:
  %val = load i32, ptr %in, align 4, !amdgpu.last.use !{}
  store i32 %val, ptr %out
  ret void
}

define amdgpu_kernel void @flat_last_use_load_1(ptr %in, ptr %out) {
; GFX12-LABEL: flat_last_use_load_1:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b128 s[0:3], s[0:1], 0x0
; GFX12-NEXT:    v_lshlrev_b32_e32 v0, 2, v0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1)
; GFX12-NEXT:    v_add_co_u32 v0, s0, s0, v0
; GFX12-NEXT:    v_add_co_ci_u32_e64 v1, null, s1, 0, s0
; GFX12-NEXT:    flat_load_b32 v2, v[0:1] th:TH_LOAD_LU
; GFX12-NEXT:    v_dual_mov_b32 v1, s3 :: v_dual_mov_b32 v0, s2
; GFX12-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-NEXT:    flat_store_b32 v[0:1], v2
; GFX12-NEXT:    s_endpgm
entry:
  %tid = call i32 @llvm.amdgcn.workitem.id.x()
  %val.gep = getelementptr inbounds i32, ptr %in, i32 %tid
  %val = load i32, ptr %val.gep, align 4, !amdgpu.last.use !{}
  store i32 %val, ptr %out
  ret void
}

define amdgpu_kernel void @flat_last_use_and_volatile_load(ptr %in, ptr %out) {
; GFX12-LABEL: flat_last_use_and_volatile_load:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b128 s[0:3], s[0:1], 0x0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX12-NEXT:    flat_load_b32 v2, v[0:1] th:TH_LOAD_BYPASS scope:SCOPE_SYS
; GFX12-NEXT:    s_wait_loadcnt 0x0
; GFX12-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX12-NEXT:    s_wait_dscnt 0x0
; GFX12-NEXT:    flat_store_b32 v[0:1], v2
; GFX12-NEXT:    s_endpgm
entry:
  %val = load volatile i32, ptr %in, align 4, !amdgpu.last.use !{}
  store i32 %val, ptr %out
  ret void
}

define amdgpu_kernel void @flat_last_use_and_nontemporal_load(ptr %in, ptr %out) {
; GFX12-LABEL: flat_last_use_and_nontemporal_load:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b128 s[0:3], s[0:1], 0x0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_dual_mov_b32 v0, s0 :: v_dual_mov_b32 v1, s1
; GFX12-NEXT:    flat_load_b32 v2, v[0:1] th:TH_LOAD_LU
; GFX12-NEXT:    v_dual_mov_b32 v0, s2 :: v_dual_mov_b32 v1, s3
; GFX12-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-NEXT:    flat_store_b32 v[0:1], v2
; GFX12-NEXT:    s_endpgm
entry:
  %val = load i32, ptr %in, align 4, !amdgpu.last.use !{}, !nontemporal !0
  store i32 %val, ptr %out
  ret void
}

!0 = !{i32 1}
declare i32 @llvm.amdgcn.workitem.id.x()
;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
; GFX12-CU: {{.*}}
; GFX12-WGP: {{.*}}