; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -amdgpu-loop-prefetch < %s | FileCheck --check-prefix=GFX12 %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -amdgpu-loop-prefetch -mattr=+safe-smem-prefetch < %s | FileCheck --check-prefix=GFX12-SPREFETCH %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx1200 -amdgpu-loop-prefetch -mattr=+safe-smem-prefetch -amdgpu-expert-scheduling-mode < %s | FileCheck --check-prefix=GFX12ES2-SPREFETCH %s
; RUN: llc -mtriple=amdgcn -mcpu=gfx1250 -amdgpu-loop-prefetch < %s | FileCheck --check-prefix=GFX1250 %s

define amdgpu_kernel void @copy_flat(ptr nocapture %d, ptr nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_flat:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB0_3
; GFX12-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-NEXT:    s_movk_i32 s4, 0xff50
; GFX12-NEXT:    s_mov_b32 s5, -1
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12-NEXT:  .LBB0_2: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    s_delay_alu instid0(SALU_CYCLE_1)
; GFX12-NEXT:    s_add_nc_u64 s[8:9], s[2:3], s[4:5]
; GFX12-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-NEXT:    v_dual_mov_b32 v5, s1 :: v_dual_mov_b32 v4, s0
; GFX12-NEXT:    v_dual_mov_b32 v0, s8 :: v_dual_mov_b32 v1, s9
; GFX12-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-NEXT:    flat_load_b128 v[0:3], v[0:1]
; GFX12-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-NEXT:    flat_store_b128 v[4:5], v[0:3]
; GFX12-NEXT:    s_cbranch_scc1 .LBB0_2
; GFX12-NEXT:  .LBB0_3: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_flat:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB0_3
; GFX12-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    s_movk_i32 s4, 0xff50
; GFX12-SPREFETCH-NEXT:    s_mov_b32 s5, -1
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12-SPREFETCH-NEXT:  .LBB0_2: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(SALU_CYCLE_1)
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[8:9], s[2:3], s[4:5]
; GFX12-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0x0, null, 0
; GFX12-SPREFETCH-NEXT:    v_dual_mov_b32 v0, s8 :: v_dual_mov_b32 v1, s9
; GFX12-SPREFETCH-NEXT:    v_dual_mov_b32 v5, s1 :: v_dual_mov_b32 v4, s0
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-SPREFETCH-NEXT:    flat_load_b128 v[0:3], v[0:1]
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-SPREFETCH-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-SPREFETCH-NEXT:    flat_store_b128 v[4:5], v[0:3]
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB0_2
; GFX12-SPREFETCH-NEXT:  .LBB0_3: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_flat:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB0_3
; GFX12ES2-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    s_movk_i32 s4, 0xff50
; GFX12ES2-SPREFETCH-NEXT:    s_mov_b32 s5, -1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12ES2-SPREFETCH-NEXT:  .LBB0_2: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(SALU_CYCLE_1)
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[8:9], s[2:3], s[4:5]
; GFX12ES2-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0x0, null, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_dual_mov_b32 v0, s8 :: v_dual_mov_b32 v1, s9
; GFX12ES2-SPREFETCH-NEXT:    v_dual_mov_b32 v5, s1 :: v_dual_mov_b32 v4, s0
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(1)
; GFX12ES2-SPREFETCH-NEXT:    flat_load_b128 v[0:3], v[0:1]
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(0)
; GFX12ES2-SPREFETCH-NEXT:    flat_store_b128 v[4:5], v[0:3]
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB0_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB0_3: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_flat:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b32 s6, s[4:5], 0x34 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s6, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB0_3
; GFX1250-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX1250-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24 nv
; GFX1250-NEXT:    v_mov_b32_e32 v0, 0
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX1250-NEXT:  .LBB0_2: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    flat_load_b128 v[2:5], v0, s[2:3] offset:-176
; GFX1250-NEXT:    flat_prefetch_b8 v0, s[2:3] scope:SCOPE_SE
; GFX1250-NEXT:    s_add_co_i32 s6, s6, -1
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX1250-NEXT:    s_cmp_lg_u32 s6, 0
; GFX1250-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX1250-NEXT:    flat_store_b128 v0, v[2:5], s[0:1]
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX1250-NEXT:    s_cbranch_scc1 .LBB0_2
; GFX1250-NEXT:  .LBB0_3: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr %s, i64 %idxprom
  %ld = load <4 x i32>, ptr %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr %d, i64 %idxprom
  store <4 x i32> %ld, ptr %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

define amdgpu_kernel void @copy_global(ptr addrspace(1) nocapture %d, ptr addrspace(1) nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_global:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB1_3
; GFX12-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-NEXT:    v_mov_b32_e32 v0, 0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12-NEXT:  .LBB1_2: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    global_load_b128 v[1:4], v0, s[2:3] offset:-176
; GFX12-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-NEXT:    s_wait_loadcnt 0x0
; GFX12-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-NEXT:    s_cbranch_scc1 .LBB1_2
; GFX12-NEXT:  .LBB1_3: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_global:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB1_3
; GFX12-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    v_mov_b32_e32 v0, 0
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12-SPREFETCH-NEXT:  .LBB1_2: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    global_load_b128 v[1:4], v0, s[2:3] offset:-176
; GFX12-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0x0, null, 0
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_wait_loadcnt 0x0
; GFX12-SPREFETCH-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB1_2
; GFX12-SPREFETCH-NEXT:  .LBB1_3: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_global:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB1_3
; GFX12ES2-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    v_mov_b32_e32 v0, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX12ES2-SPREFETCH-NEXT:  .LBB1_2: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(0) depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    global_load_b128 v[1:4], v0, s[2:3] offset:-176
; GFX12ES2-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0x0, null, 0
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_loadcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB1_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB1_3: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_global:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b32 s6, s[4:5], 0x34 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s6, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB1_3
; GFX1250-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX1250-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24 nv
; GFX1250-NEXT:    v_mov_b32_e32 v0, 0
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 0xb0
; GFX1250-NEXT:  .LBB1_2: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    global_load_b128 v[2:5], v0, s[2:3] offset:-176
; GFX1250-NEXT:    global_prefetch_b8 v0, s[2:3] scope:SCOPE_SE
; GFX1250-NEXT:    s_add_co_i32 s6, s6, -1
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX1250-NEXT:    s_cmp_lg_u32 s6, 0
; GFX1250-NEXT:    s_wait_loadcnt 0x0
; GFX1250-NEXT:    global_store_b128 v0, v[2:5], s[0:1]
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX1250-NEXT:    s_cbranch_scc1 .LBB1_2
; GFX1250-NEXT:  .LBB1_3: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr addrspace(1) %s, i64 %idxprom
  %ld = load <4 x i32>, ptr addrspace(1) %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr addrspace(1) %d, i64 %idxprom
  store <4 x i32> %ld, ptr addrspace(1) %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

define amdgpu_kernel void @copy_constant(ptr addrspace(1) nocapture %d, ptr addrspace(4) nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_constant:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB2_3
; GFX12-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-NEXT:    v_mov_b32_e32 v0, 0
; GFX12-NEXT:  .LBB2_2: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_load_b128 s[8:11], s[2:3], 0x0
; GFX12-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_dual_mov_b32 v1, s8 :: v_dual_mov_b32 v2, s9
; GFX12-NEXT:    v_dual_mov_b32 v3, s10 :: v_dual_mov_b32 v4, s11
; GFX12-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-NEXT:    s_cbranch_scc1 .LBB2_2
; GFX12-NEXT:  .LBB2_3: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_constant:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB2_3
; GFX12-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    v_mov_b32_e32 v0, 0
; GFX12-SPREFETCH-NEXT:  .LBB2_2: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[8:11], s[2:3], 0x0
; GFX12-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0xb0, null, 0
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    v_dual_mov_b32 v1, s8 :: v_dual_mov_b32 v2, s9
; GFX12-SPREFETCH-NEXT:    v_dual_mov_b32 v3, s10 :: v_dual_mov_b32 v4, s11
; GFX12-SPREFETCH-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB2_2
; GFX12-SPREFETCH-NEXT:  .LBB2_3: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_constant:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b32 s6, s[4:5], 0x34
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB2_3
; GFX12ES2-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    v_mov_b32_e32 v0, 0
; GFX12ES2-SPREFETCH-NEXT:  .LBB2_2: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[8:11], s[2:3], 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_prefetch_data s[2:3], 0xb0, null, 0
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s6, s6, -1
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s6, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_dual_mov_b32 v1, s8 :: v_dual_mov_b32 v2, s9
; GFX12ES2-SPREFETCH-NEXT:    v_dual_mov_b32 v3, s10 :: v_dual_mov_b32 v4, s11
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(0)
; GFX12ES2-SPREFETCH-NEXT:    global_store_b128 v0, v[1:4], s[0:1]
; GFX12ES2-SPREFETCH-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB2_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB2_3: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_constant:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b32 s6, s[4:5], 0x34 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s6, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB2_3
; GFX1250-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX1250-NEXT:    s_load_b128 s[0:3], s[4:5], 0x24 nv
; GFX1250-NEXT:    v_mov_b32_e32 v0, 0
; GFX1250-NEXT:  .LBB2_2: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    global_prefetch_b8 v0, s[2:3] offset:176 scope:SCOPE_SE
; GFX1250-NEXT:    s_load_b128 s[8:11], s[2:3], 0x0 nv
; GFX1250-NEXT:    s_add_co_i32 s6, s6, -1
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[2:3], s[2:3], 16
; GFX1250-NEXT:    s_cmp_lg_u32 s6, 0
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    v_mov_b64_e32 v[2:3], s[8:9]
; GFX1250-NEXT:    v_mov_b64_e32 v[4:5], s[10:11]
; GFX1250-NEXT:    global_store_b128 v0, v[2:5], s[0:1]
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    s_add_nc_u64 s[0:1], s[0:1], 16
; GFX1250-NEXT:    s_cbranch_scc1 .LBB2_2
; GFX1250-NEXT:  .LBB2_3: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr addrspace(4) %s, i64 %idxprom
  %ld = load <4 x i32>, ptr addrspace(4) %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr addrspace(1) %d, i64 %idxprom
  store <4 x i32> %ld, ptr addrspace(1) %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

define amdgpu_kernel void @copy_local(ptr addrspace(3) nocapture %d, ptr addrspace(3) nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_local:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b96 s[0:2], s[4:5], 0x24
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s2, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB3_2
; GFX12-NEXT:  .LBB3_1: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-NEXT:    v_mov_b32_e32 v2, s1
; GFX12-NEXT:    v_mov_b32_e32 v4, s0
; GFX12-NEXT:    s_add_co_i32 s2, s2, -1
; GFX12-NEXT:    s_add_co_i32 s0, s0, 16
; GFX12-NEXT:    s_add_co_i32 s1, s1, 16
; GFX12-NEXT:    ds_load_2addr_b32 v[0:1], v2 offset0:2 offset1:3
; GFX12-NEXT:    ds_load_2addr_b32 v[2:3], v2 offset1:1
; GFX12-NEXT:    s_cmp_lg_u32 s2, 0
; GFX12-NEXT:    s_wait_dscnt 0x1
; GFX12-NEXT:    ds_store_2addr_b32 v4, v0, v1 offset0:2 offset1:3
; GFX12-NEXT:    s_wait_dscnt 0x1
; GFX12-NEXT:    ds_store_2addr_b32 v4, v2, v3 offset1:1
; GFX12-NEXT:    s_cbranch_scc1 .LBB3_1
; GFX12-NEXT:  .LBB3_2: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_local:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b96 s[0:2], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s2, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB3_2
; GFX12-SPREFETCH-NEXT:  .LBB3_1: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-SPREFETCH-NEXT:    v_mov_b32_e32 v2, s1
; GFX12-SPREFETCH-NEXT:    v_mov_b32_e32 v4, s0
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s2, s2, -1
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, 16
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s1, s1, 16
; GFX12-SPREFETCH-NEXT:    ds_load_2addr_b32 v[0:1], v2 offset0:2 offset1:3
; GFX12-SPREFETCH-NEXT:    ds_load_2addr_b32 v[2:3], v2 offset1:1
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s2, 0
; GFX12-SPREFETCH-NEXT:    s_wait_dscnt 0x1
; GFX12-SPREFETCH-NEXT:    ds_store_2addr_b32 v4, v0, v1 offset0:2 offset1:3
; GFX12-SPREFETCH-NEXT:    s_wait_dscnt 0x1
; GFX12-SPREFETCH-NEXT:    ds_store_2addr_b32 v4, v2, v3 offset1:1
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB3_1
; GFX12-SPREFETCH-NEXT:  .LBB3_2: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_local:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b96 s[0:2], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s2, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB3_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB3_1: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0) depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_mov_b32_e32 v2, s1
; GFX12ES2-SPREFETCH-NEXT:    v_mov_b32_e32 v4, s0
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s2, s2, -1
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, 16
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s1, s1, 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(1)
; GFX12ES2-SPREFETCH-NEXT:    ds_load_2addr_b32 v[0:1], v2 offset0:2 offset1:3
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    ds_load_2addr_b32 v[2:3], v2 offset1:1
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s2, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_dscnt 0x1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(0)
; GFX12ES2-SPREFETCH-NEXT:    ds_store_2addr_b32 v4, v0, v1 offset0:2 offset1:3
; GFX12ES2-SPREFETCH-NEXT:    s_wait_dscnt 0x1
; GFX12ES2-SPREFETCH-NEXT:    ds_store_2addr_b32 v4, v2, v3 offset1:1
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB3_1
; GFX12ES2-SPREFETCH-NEXT:  .LBB3_2: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_local:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b96 s[0:2], s[4:5], 0x24 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s2, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB3_2
; GFX1250-NEXT:  .LBB3_1: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    v_dual_mov_b32 v2, s1 :: v_dual_mov_b32 v4, s0
; GFX1250-NEXT:    s_add_co_i32 s2, s2, -1
; GFX1250-NEXT:    s_add_co_i32 s0, s0, 16
; GFX1250-NEXT:    s_add_co_i32 s1, s1, 16
; GFX1250-NEXT:    ds_load_2addr_b32 v[0:1], v2 offset0:2 offset1:3
; GFX1250-NEXT:    ds_load_2addr_b32 v[2:3], v2 offset1:1
; GFX1250-NEXT:    s_cmp_lg_u32 s2, 0
; GFX1250-NEXT:    s_wait_dscnt 0x1
; GFX1250-NEXT:    ds_store_2addr_b32 v4, v0, v1 offset0:2 offset1:3
; GFX1250-NEXT:    s_wait_dscnt 0x1
; GFX1250-NEXT:    ds_store_2addr_b32 v4, v2, v3 offset1:1
; GFX1250-NEXT:    s_cbranch_scc1 .LBB3_1
; GFX1250-NEXT:  .LBB3_2: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr addrspace(3) %s, i64 %idxprom
  %ld = load <4 x i32>, ptr addrspace(3) %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr addrspace(3) %d, i64 %idxprom
  store <4 x i32> %ld, ptr addrspace(3) %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

define amdgpu_kernel void @copy_flat_divergent(ptr nocapture %d, ptr nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_flat_divergent:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB4_3
; GFX12-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-NEXT:  .LBB4_2: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_2)
; GFX12-NEXT:    v_add_co_u32 v4, vcc_lo, 0xffffff50, v2
; GFX12-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v5, null, -1, v3, vcc_lo
; GFX12-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-NEXT:    flat_load_b128 v[4:7], v[4:5]
; GFX12-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-NEXT:    flat_store_b128 v[0:1], v[4:7]
; GFX12-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12-NEXT:    s_cbranch_scc1 .LBB4_2
; GFX12-NEXT:  .LBB4_3: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_flat_divergent:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB4_3
; GFX12-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-SPREFETCH-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-SPREFETCH-NEXT:  .LBB4_2: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_2)
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v4, vcc_lo, 0xffffff50, v2
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v5, null, -1, v3, vcc_lo
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-SPREFETCH-NEXT:    flat_load_b128 v[4:7], v[4:5]
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12-SPREFETCH-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12-SPREFETCH-NEXT:    flat_store_b128 v[0:1], v[4:7]
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB4_2
; GFX12-SPREFETCH-NEXT:  .LBB4_3: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_flat_divergent:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB4_3
; GFX12ES2-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12ES2-SPREFETCH-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:  .LBB4_2: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_2)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v4, vcc_lo, 0xffffff50, v2
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v5, null, -1, v3, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(2)
; GFX12ES2-SPREFETCH-NEXT:    flat_load_b128 v[4:7], v[4:5]
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    flat_store_b128 v[0:1], v[4:7]
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB4_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB4_3: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_flat_divergent:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b32 s2, s[4:5], 0x34 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s2, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB4_3
; GFX1250-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX1250-NEXT:    s_load_b128 s[8:11], s[4:5], 0x24 nv
; GFX1250-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX1250-NEXT:    v_mov_b32_e32 v1, 0
; GFX1250-NEXT:    s_mov_b64 s[0:1], 0xffffffffffffff50
; GFX1250-NEXT:    s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX1250-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], s[10:11], v[0:1]
; GFX1250-NEXT:    v_add_nc_u64_e32 v[0:1], s[8:9], v[0:1]
; GFX1250-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], 0xb0, v[2:3]
; GFX1250-NEXT:  .LBB4_2: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_3) | instid1(SALU_CYCLE_1)
; GFX1250-NEXT:    v_add_nc_u64_e32 v[4:5], s[0:1], v[2:3]
; GFX1250-NEXT:    flat_prefetch_b8 v[2:3] scope:SCOPE_SE
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], 16, v[2:3]
; GFX1250-NEXT:    s_add_co_i32 s2, s2, -1
; GFX1250-NEXT:    s_cmp_lg_u32 s2, 0
; GFX1250-NEXT:    flat_load_b128 v[4:7], v[4:5]
; GFX1250-NEXT:    s_wait_loadcnt_dscnt 0x0
; GFX1250-NEXT:    flat_store_b128 v[0:1], v[4:7]
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    v_add_nc_u64_e32 v[0:1], 16, v[0:1]
; GFX1250-NEXT:    s_cbranch_scc1 .LBB4_2
; GFX1250-NEXT:  .LBB4_3: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %tid = call i32 @llvm.amdgcn.workitem.id.x()
  %s.tid = getelementptr inbounds <4 x i32>, ptr %s, i32 %tid
  %d.tid = getelementptr inbounds <4 x i32>, ptr %d, i32 %tid
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr %s.tid, i64 %idxprom
  %ld = load <4 x i32>, ptr %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr %d.tid, i64 %idxprom
  store <4 x i32> %ld, ptr %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

define amdgpu_kernel void @copy_global_divergent(ptr addrspace(1) nocapture %d, ptr addrspace(1) nocapture readonly %s, i32 %n) {
; GFX12-LABEL: copy_global_divergent:
; GFX12:       ; %bb.0: ; %entry
; GFX12-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12-NEXT:    s_cbranch_scc1 .LBB5_3
; GFX12-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12-NEXT:    s_wait_kmcnt 0x0
; GFX12-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-NEXT:  .LBB5_2: ; %for.body
; GFX12-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-NEXT:    global_load_b128 v[4:7], v[2:3], off offset:-176
; GFX12-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12-NEXT:    s_wait_loadcnt 0x0
; GFX12-NEXT:    global_store_b128 v[0:1], v[4:7], off
; GFX12-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12-NEXT:    s_cbranch_scc1 .LBB5_2
; GFX12-NEXT:  .LBB5_3: ; %for.end
; GFX12-NEXT:    s_endpgm
;
; GFX12-SPREFETCH-LABEL: copy_global_divergent:
; GFX12-SPREFETCH:       ; %bb.0: ; %entry
; GFX12-SPREFETCH-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB5_3
; GFX12-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12-SPREFETCH-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12-SPREFETCH-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-SPREFETCH-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-SPREFETCH-NEXT:  .LBB5_2: ; %for.body
; GFX12-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12-SPREFETCH-NEXT:    global_load_b128 v[4:7], v[2:3], off offset:-176
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12-SPREFETCH-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12-SPREFETCH-NEXT:    s_wait_loadcnt 0x0
; GFX12-SPREFETCH-NEXT:    global_store_b128 v[0:1], v[4:7], off
; GFX12-SPREFETCH-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB5_2
; GFX12-SPREFETCH-NEXT:  .LBB5_3: ; %for.end
; GFX12-SPREFETCH-NEXT:    s_endpgm
;
; GFX12ES2-SPREFETCH-LABEL: copy_global_divergent:
; GFX12ES2-SPREFETCH:       ; %bb.0: ; %entry
; GFX12ES2-SPREFETCH-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_SCHED_MODE, 0, 2), 2
; GFX12ES2-SPREFETCH-NEXT:    s_load_b32 s0, s[4:5], 0x34
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_eq_u32 s0, 0
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB5_3
; GFX12ES2-SPREFETCH-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX12ES2-SPREFETCH-NEXT:    s_load_b128 s[4:7], s[4:5], 0x24
; GFX12ES2-SPREFETCH-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12ES2-SPREFETCH-NEXT:    v_lshlrev_b32_e32 v0, 4, v0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_kmcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, s1, s6, v0
; GFX12ES2-SPREFETCH-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v0, s1, s4, v0
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_sdst(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, s5, 0, s1
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:  .LBB5_2: ; %for.body
; GFX12ES2-SPREFETCH-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vdst(0)
; GFX12ES2-SPREFETCH-NEXT:    global_load_b128 v[4:7], v[2:3], off offset:-176
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v2, vcc_lo, v2, 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v3, null, 0, v3, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:    s_add_co_i32 s0, s0, -1
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_sa_sdst(0)
; GFX12ES2-SPREFETCH-NEXT:    s_cmp_lg_u32 s0, 0
; GFX12ES2-SPREFETCH-NEXT:    s_wait_loadcnt 0x0
; GFX12ES2-SPREFETCH-NEXT:    global_store_b128 v[0:1], v[4:7], off
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_vm_vsrc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_u32 v0, vcc_lo, v0, 16
; GFX12ES2-SPREFETCH-NEXT:    s_wait_alu depctr_va_vcc(0)
; GFX12ES2-SPREFETCH-NEXT:    v_add_co_ci_u32_e64 v1, null, 0, v1, vcc_lo
; GFX12ES2-SPREFETCH-NEXT:    s_cbranch_scc1 .LBB5_2
; GFX12ES2-SPREFETCH-NEXT:  .LBB5_3: ; %for.end
; GFX12ES2-SPREFETCH-NEXT:    s_endpgm
;
; GFX1250-LABEL: copy_global_divergent:
; GFX1250:       ; %bb.0: ; %entry
; GFX1250-NEXT:    s_setreg_imm32_b32 hwreg(HW_REG_WAVE_MODE, 25, 1), 1 ; msbs: dst=0 src0=0 src1=0 src2=0
; GFX1250-NEXT:    s_load_b32 s0, s[4:5], 0x34 nv
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    s_cmp_eq_u32 s0, 0
; GFX1250-NEXT:    s_cbranch_scc1 .LBB5_3
; GFX1250-NEXT:  ; %bb.1: ; %for.body.preheader
; GFX1250-NEXT:    s_load_b128 s[8:11], s[4:5], 0x24 nv
; GFX1250-NEXT:    v_and_b32_e32 v0, 0x3ff, v0
; GFX1250-NEXT:    s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX1250-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_lshlrev_b32 v0, 4, v0
; GFX1250-NEXT:    s_wait_kmcnt 0x0
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], s[10:11], v[0:1]
; GFX1250-NEXT:    v_add_nc_u64_e32 v[0:1], s[8:9], v[0:1]
; GFX1250-NEXT:    s_delay_alu instid0(VALU_DEP_2)
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], 0xb0, v[2:3]
; GFX1250-NEXT:  .LBB5_2: ; %for.body
; GFX1250-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX1250-NEXT:    global_load_b128 v[4:7], v[2:3], off offset:-176
; GFX1250-NEXT:    global_prefetch_b8 v[2:3], off scope:SCOPE_SE
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    v_add_nc_u64_e32 v[2:3], 16, v[2:3]
; GFX1250-NEXT:    s_add_co_i32 s0, s0, -1
; GFX1250-NEXT:    s_delay_alu instid0(SALU_CYCLE_1)
; GFX1250-NEXT:    s_cmp_lg_u32 s0, 0
; GFX1250-NEXT:    s_wait_loadcnt 0x0
; GFX1250-NEXT:    global_store_b128 v[0:1], v[4:7], off
; GFX1250-NEXT:    s_wait_xcnt 0x0
; GFX1250-NEXT:    v_add_nc_u64_e32 v[0:1], 16, v[0:1]
; GFX1250-NEXT:    s_cbranch_scc1 .LBB5_2
; GFX1250-NEXT:  .LBB5_3: ; %for.end
; GFX1250-NEXT:    s_endpgm
entry:
  %tid = call i32 @llvm.amdgcn.workitem.id.x()
  %s.tid = getelementptr inbounds <4 x i32>, ptr addrspace(1) %s, i32 %tid
  %d.tid = getelementptr inbounds <4 x i32>, ptr addrspace(1) %d, i32 %tid
  %cmp6.not = icmp eq i32 %n, 0
  br i1 %cmp6.not, label %for.end, label %for.body

for.body:                                         ; preds = %entry, %for.body
  %i.07 = phi i32 [ %inc, %for.body ], [ 0, %entry ]
  %idxprom = zext i32 %i.07 to i64
  %arrayidx = getelementptr inbounds <4 x i32>, ptr addrspace(1) %s.tid, i64 %idxprom
  %ld = load <4 x i32>, ptr addrspace(1) %arrayidx, align 4
  %arrayidx2 = getelementptr inbounds <4 x i32>, ptr addrspace(1) %d.tid, i64 %idxprom
  store <4 x i32> %ld, ptr addrspace(1) %arrayidx2, align 4
  %inc = add nuw i32 %i.07, 1
  %exitcond.not = icmp eq i32 %inc, %n
  br i1 %exitcond.not, label %for.end, label %for.body

for.end:                                          ; preds = %for.body, %entry
  ret void
}

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