; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
; RUN: llc -O0 -mtriple=amdgcn- -mcpu=gfx900 -amdgpu-dpp-combine=false -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9-O0 %s
; RUN: llc -mtriple=amdgcn- -mcpu=gfx900 -amdgpu-dpp-combine=false -verify-machineinstrs < %s | FileCheck -check-prefix=GFX9-O3 %s

; NOTE: llvm.amdgcn.wwm is deprecated, use llvm.amdgcn.strict.wwm instead.

define amdgpu_cs void @no_cfg(ptr addrspace(8) inreg %tmp14) {
; GFX9-O0-LABEL: no_cfg:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s6, s2
; GFX9-O0-NEXT:    s_mov_b32 s4, s0
; GFX9-O0-NEXT:    ; kill: def $sgpr6 killed $sgpr6 def $sgpr6_sgpr7
; GFX9-O0-NEXT:    s_mov_b32 s7, s3
; GFX9-O0-NEXT:    s_mov_b32 s8, s7
; GFX9-O0-NEXT:    s_mov_b32 s9, s6
; GFX9-O0-NEXT:    ; kill: def $sgpr4 killed $sgpr4 def $sgpr4_sgpr5
; GFX9-O0-NEXT:    s_mov_b32 s5, s1
; GFX9-O0-NEXT:    s_mov_b32 s10, s5
; GFX9-O0-NEXT:    s_mov_b32 s0, s4
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s10
; GFX9-O0-NEXT:    s_mov_b32 s2, s9
; GFX9-O0-NEXT:    s_mov_b32 s3, s8
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[5:6], off, s[0:3], s4
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v6
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, s4
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s4
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s4
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v2, v0 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v0, v0, v2
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, s4
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v0, v1 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v0, v1, v0
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v0
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[6:7], v3, v4
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v3, 0, 1, s[6:7]
; GFX9-O0-NEXT:    s_mov_b32 s5, 1
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s5, v3
; GFX9-O0-NEXT:    s_mov_b32 s5, 2
; GFX9-O0-NEXT:    v_and_b32_e64 v3, v3, s5
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: no_cfg:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[4:5], off, s[0:3], 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v4, s[4:5]
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v3, 0, v5, s[4:5]
; GFX9-O3-NEXT:    s_nop 0
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_mov_b32_dpp v0, v3 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    v_add_u32_e32 v0, v3, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, v4, v5
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v4, 0, 1, vcc
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 1, v4
; GFX9-O3-NEXT:    v_and_b32_e32 v4, 2, v4
; GFX9-O3-NEXT:    buffer_store_dword v4, off, s[0:3], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm
  %tmp100 = call <2 x float> @llvm.amdgcn.raw.ptr.buffer.load.v2f32(ptr addrspace(8) %tmp14, i32 0, i32 0, i32 0)
  %tmp101 = bitcast <2 x float> %tmp100 to <2 x i32>
  %tmp102 = extractelement <2 x i32> %tmp101, i32 0
  %tmp103 = extractelement <2 x i32> %tmp101, i32 1
  %tmp105 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp102, i32 0)
  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp103, i32 0)


  %tmp120 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp105, i32 323, i32 12, i32 15, i1 false)
  %tmp121 = add i32 %tmp105, %tmp120
  %tmp122 = tail call i32 @llvm.amdgcn.wwm.i32(i32 %tmp121)

  %tmp135 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp107, i32 323, i32 12, i32 15, i1 false)
  %tmp136 = add i32 %tmp107, %tmp135
  %tmp137 = tail call i32 @llvm.amdgcn.wwm.i32(i32 %tmp136)

  %tmp138 = icmp eq i32 %tmp122, %tmp137
  %tmp139 = sext i1 %tmp138 to i32
  %tmp140 = shl nsw i32 %tmp139, 1
  %tmp141 = and i32 %tmp140, 2
  %tmp145 = bitcast i32 %tmp141 to float
  call void @llvm.amdgcn.raw.ptr.buffer.store.f32(float %tmp145, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define amdgpu_cs void @cfg(ptr addrspace(8) inreg %tmp14, i32 %arg) {
; GFX9-O0-LABEL: cfg:
; GFX9-O0:       ; %bb.0: ; %entry
; GFX9-O0-NEXT:    s_mov_b32 s16, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s17, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s18, -1
; GFX9-O0-NEXT:    s_mov_b32 s19, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s16, s16, s4
; GFX9-O0-NEXT:    s_addc_u32 s17, s17, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr3 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[16:19], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 0
; GFX9-O0-NEXT:    s_mov_b32 s4, s1
; GFX9-O0-NEXT:    v_readlane_b32 s1, v0, 0
; GFX9-O0-NEXT:    ; kill: def $sgpr2 killed $sgpr2 def $sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s3, s1
; GFX9-O0-NEXT:    s_mov_b32 s8, s3
; GFX9-O0-NEXT:    s_mov_b32 s9, s2
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1
; GFX9-O0-NEXT:    s_mov_b32 s1, s4
; GFX9-O0-NEXT:    s_mov_b32 s10, s1
; GFX9-O0-NEXT:    s_mov_b32 s4, s0
; GFX9-O0-NEXT:    ; kill: def $sgpr4 killed $sgpr4 def $sgpr4_sgpr5_sgpr6_sgpr7
; GFX9-O0-NEXT:    s_mov_b32 s5, s10
; GFX9-O0-NEXT:    s_mov_b32 s6, s9
; GFX9-O0-NEXT:    s_mov_b32 s7, s8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s2, 1
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 2
; GFX9-O0-NEXT:    v_writelane_b32 v0, s0, 3
; GFX9-O0-NEXT:    v_writelane_b32 v0, s1, 4
; GFX9-O0-NEXT:    s_mov_b32 s0, 0
; GFX9-O0-NEXT:    s_nop 2
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[4:5], off, s[4:7], s0
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    buffer_store_dword v4, off, s[16:19], 0 offset:12 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    buffer_store_dword v5, off, s[16:19], 0 offset:16 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    ; implicit-def: $sgpr2_sgpr3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[2:3], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s0
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v2, v1 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v1, v2
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    buffer_store_dword v4, off, s[16:19], 0 offset:8 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[2:3], v3, s0
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, s0
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[16:19], 0 offset:4 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], exec
; GFX9-O0-NEXT:    v_writelane_b32 v0, s0, 5
; GFX9-O0-NEXT:    v_writelane_b32 v0, s1, 6
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[16:19], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_and_b64 s[0:1], s[0:1], s[2:3]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[0:1]
; GFX9-O0-NEXT:    s_cbranch_execz .LBB1_2
; GFX9-O0-NEXT:  ; %bb.1: ; %if
; GFX9-O0-NEXT:    buffer_load_dword v3, off, s[16:19], 0 offset:12 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    buffer_load_dword v4, off, s[16:19], 0 offset:16 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[0:1], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v2, v1, v0, s[0:1]
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v2, v1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[16:19], 0 offset:4 ; 4-byte Folded Spill
; GFX9-O0-NEXT:  .LBB1_2: ; %merge
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[16:19], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s4, v0, 5
; GFX9-O0-NEXT:    v_readlane_b32 s5, v0, 6
; GFX9-O0-NEXT:    s_or_b64 exec, exec, s[4:5]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v0, 1
; GFX9-O0-NEXT:    v_readlane_b32 s3, v0, 2
; GFX9-O0-NEXT:    v_readlane_b32 s0, v0, 3
; GFX9-O0-NEXT:    v_readlane_b32 s1, v0, 4
; GFX9-O0-NEXT:    buffer_load_dword v3, off, s[16:19], 0 offset:8 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    buffer_load_dword v4, off, s[16:19], 0 offset:4 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[4:5], v3, v4
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v3, 0, 1, s[4:5]
; GFX9-O0-NEXT:    s_mov_b32 s4, 1
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s4, v3
; GFX9-O0-NEXT:    s_mov_b32 s4, 2
; GFX9-O0-NEXT:    v_and_b32_e64 v3, v3, s4
; GFX9-O0-NEXT:    s_mov_b32 s6, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 killed $sgpr0_sgpr1
; GFX9-O0-NEXT:    s_mov_b32 s4, s3
; GFX9-O0-NEXT:    s_mov_b32 s5, s2
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s6
; GFX9-O0-NEXT:    s_mov_b32 s2, s5
; GFX9-O0-NEXT:    s_mov_b32 s3, s4
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: cfg:
; GFX9-O3:       ; %bb.0: ; %entry
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[3:4], off, s[0:3], 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v3, s[4:5]
; GFX9-O3-NEXT:    s_nop 1
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v0
; GFX9-O3-NEXT:    s_and_saveexec_b64 s[4:5], vcc
; GFX9-O3-NEXT:  ; %bb.1: ; %if
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v4, s[6:7]
; GFX9-O3-NEXT:    s_nop 1
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v1
; GFX9-O3-NEXT:  ; %bb.2: ; %merge
; GFX9-O3-NEXT:    s_or_b64 exec, exec, s[4:5]
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, v3, v5
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v0, 0, 1, vcc
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v0, 1, v0
; GFX9-O3-NEXT:    v_and_b32_e32 v0, 2, v0
; GFX9-O3-NEXT:    buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm
entry:
  %tmp100 = call <2 x float> @llvm.amdgcn.raw.ptr.buffer.load.v2f32(ptr addrspace(8) %tmp14, i32 0, i32 0, i32 0)
  %tmp101 = bitcast <2 x float> %tmp100 to <2 x i32>
  %tmp102 = extractelement <2 x i32> %tmp101, i32 0
  %tmp105 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp102, i32 0)

  %tmp120 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp105, i32 323, i32 12, i32 15, i1 false)
  %tmp121 = add i32 %tmp105, %tmp120
  %tmp122 = tail call i32 @llvm.amdgcn.wwm.i32(i32 %tmp121)

  %cond = icmp eq i32 %arg, 0
  br i1 %cond, label %if, label %merge
if:
  %tmp103 = extractelement <2 x i32> %tmp101, i32 1
  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp103, i32 0)

  %tmp135 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp107, i32 323, i32 12, i32 15, i1 false)
  %tmp136 = add i32 %tmp107, %tmp135
  %tmp137 = tail call i32 @llvm.amdgcn.wwm.i32(i32 %tmp136)
  br label %merge

merge:
  %merge_value = phi i32 [ 0, %entry ], [%tmp137, %if ]
  %tmp138 = icmp eq i32 %tmp122, %merge_value
  %tmp139 = sext i1 %tmp138 to i32
  %tmp140 = shl nsw i32 %tmp139, 1
  %tmp141 = and i32 %tmp140, 2
  %tmp145 = bitcast i32 %tmp141 to float
  call void @llvm.amdgcn.raw.ptr.buffer.store.f32(float %tmp145, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define hidden i32 @called(i32 %a) noinline {
; GFX9-O0-LABEL: called:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v0, v0
; GFX9-O0-NEXT:    v_mul_lo_u32 v0, v1, v0
; GFX9-O0-NEXT:    v_sub_u32_e64 v0, v0, v1
; GFX9-O0-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-O3-LABEL: called:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v0, v0
; GFX9-O3-NEXT:    v_mul_lo_u32 v0, v1, v0
; GFX9-O3-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-O3-NEXT:    s_setpc_b64 s[30:31]
  %add = add i32 %a, %a
  %mul = mul i32 %add, %a
  %sub = sub i32 %mul, %add
  ret i32 %sub
}

define amdgpu_kernel void @call(ptr addrspace(8) inreg %tmp14, i32 inreg %arg) {
; GFX9-O0-LABEL: call:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s32, 0x400
; GFX9-O0-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s26, -1
; GFX9-O0-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O0-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr7 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[10:11], -1
; GFX9-O0-NEXT:    v_writelane_b32 v7, s10, 0
; GFX9-O0-NEXT:    v_writelane_b32 v7, s11, 1
; GFX9-O0-NEXT:    s_mov_b32 s14, s8
; GFX9-O0-NEXT:    s_mov_b32 s13, s7
; GFX9-O0-NEXT:    s_mov_b32 s12, s6
; GFX9-O0-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[2:3]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v7, 0
; GFX9-O0-NEXT:    v_readlane_b32 s3, v7, 1
; GFX9-O0-NEXT:    v_writelane_b32 v7, s4, 2
; GFX9-O0-NEXT:    v_writelane_b32 v7, s5, 3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v7, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O0-NEXT:    v_readlane_b32 s0, v7, 2
; GFX9-O0-NEXT:    v_readlane_b32 s1, v7, 3
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x24
; GFX9-O0-NEXT:    s_load_dwordx2 s[8:9], s[0:1], 0x2c
; GFX9-O0-NEXT:    s_load_dword s2, s[0:1], 0x34
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s6, 4
; GFX9-O0-NEXT:    v_writelane_b32 v0, s7, 5
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_mov_b32 s3, s9
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 killed $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 killed $sgpr16_sgpr17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 def $sgpr16_sgpr17_sgpr18_sgpr19
; GFX9-O0-NEXT:    s_mov_b32 s17, s9
; GFX9-O0-NEXT:    s_mov_b32 s18, s8
; GFX9-O0-NEXT:    s_mov_b32 s19, s3
; GFX9-O0-NEXT:    v_writelane_b32 v0, s16, 6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s17, 7
; GFX9-O0-NEXT:    v_writelane_b32 v0, s18, 8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s19, 9
; GFX9-O0-NEXT:    s_mov_b32 s3, 0
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 10
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, s3
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, s2
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 56
; GFX9-O0-NEXT:    s_mov_b32 s2, s0
; GFX9-O0-NEXT:    s_mov_b32 s0, s1
; GFX9-O0-NEXT:    s_mov_b32 s3, s6
; GFX9-O0-NEXT:    s_mov_b32 s1, s7
; GFX9-O0-NEXT:    s_add_u32 s8, s2, s3
; GFX9-O0-NEXT:    s_addc_u32 s0, s0, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s0
; GFX9-O0-NEXT:    s_getpc_b64 s[16:17]
; GFX9-O0-NEXT:    s_add_u32 s16, s16, called@rel32@lo+4
; GFX9-O0-NEXT:    s_addc_u32 s17, s17, called@rel32@hi+12
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O0-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O0-NEXT:    s_mov_b32 s6, 20
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s6, v3
; GFX9-O0-NEXT:    s_mov_b32 s6, 10
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v4, s6, v4
; GFX9-O0-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6_sgpr7
; GFX9-O0-NEXT:    ; implicit-def: $sgpr15
; GFX9-O0-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O0-NEXT:    s_swappc_b64 s[30:31], s[16:17]
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v1, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s0, v1, 6
; GFX9-O0-NEXT:    v_readlane_b32 s1, v1, 7
; GFX9-O0-NEXT:    v_readlane_b32 s2, v1, 8
; GFX9-O0-NEXT:    v_readlane_b32 s3, v1, 9
; GFX9-O0-NEXT:    v_readlane_b32 s6, v1, 4
; GFX9-O0-NEXT:    v_readlane_b32 s7, v1, 5
; GFX9-O0-NEXT:    v_readlane_b32 s4, v1, 10
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_add_u32_e64 v3, v3, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    buffer_store_dword v1, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: call:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O3-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O3-NEXT:    s_mov_b32 s26, -1
; GFX9-O3-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O3-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O3-NEXT:    s_mov_b32 s32, 0
; GFX9-O3-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O3-NEXT:    s_mov_b32 s14, s8
; GFX9-O3-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O3-NEXT:    s_load_dword s4, s[2:3], 0x34
; GFX9-O3-NEXT:    s_load_dwordx4 s[16:19], s[2:3], 0x24
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, 0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s4
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_add_u32 s8, s2, 56
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v3, 20, v3
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 10, v4
; GFX9-O3-NEXT:    s_addc_u32 s9, s3, 0
; GFX9-O3-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O3-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O3-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O3-NEXT:    s_mov_b32 s12, s6
; GFX9-O3-NEXT:    s_mov_b32 s13, s7
; GFX9-O3-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O3-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O3-NEXT:    s_getpc_b64 s[22:23]
; GFX9-O3-NEXT:    s_add_u32 s22, s22, called@rel32@lo+4
; GFX9-O3-NEXT:    s_addc_u32 s23, s23, called@rel32@hi+12
; GFX9-O3-NEXT:    s_swappc_b64 s[30:31], s[22:23]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O3-NEXT:    v_add_u32_e32 v3, v3, v6
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O3-NEXT:    buffer_store_dword v0, off, s[16:19], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm


  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %arg, i32 0)
  %tmp134 = call i32 @called(i32 %tmp107)
  %tmp136 = add i32 %tmp134, %tmp107
  %tmp137 = tail call i32 @llvm.amdgcn.wwm.i32(i32 %tmp136)
  call void @llvm.amdgcn.raw.ptr.buffer.store.i32(i32 %tmp137, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define i64 @called_i64(i64 %a) noinline {
; GFX9-O0-LABEL: called_i64:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v0
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4
; GFX9-O0-NEXT:    ; kill: def $vgpr2 killed $vgpr2 def $vgpr2_vgpr3 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4_sgpr5
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_add_co_u32_e64 v4, s[4:5], v4, v5
; GFX9-O0-NEXT:    v_addc_co_u32_e64 v0, s[4:5], v0, v1, s[4:5]
; GFX9-O0-NEXT:    ; kill: def $vgpr4 killed $vgpr4 def $vgpr4_vgpr5 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_mov_b32 s4, 32
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_lshrrev_b64 v[0:1], s4, v[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    v_mul_lo_u32 v1, v0, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_lshrrev_b64 v[6:7], s4, v[4:5]
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    v_mul_lo_u32 v2, v2, v3
; GFX9-O0-NEXT:    v_mad_u64_u32 v[6:7], s[6:7], v0, v3, 0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_add3_u32 v0, v0, v1, v2
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s5
; GFX9-O0-NEXT:    ; kill: def $vgpr0 killed $vgpr0 def $vgpr0_vgpr1 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v2
; GFX9-O0-NEXT:    v_lshlrev_b64 v[1:2], s4, v[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 killed $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    s_mov_b32 s5, 0
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, 0
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 def $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_or_b32_e64 v0, v0, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v6
; GFX9-O0-NEXT:    v_or_b32_e64 v6, v1, v2
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 def $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v6
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v4
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v5
; GFX9-O0-NEXT:    v_sub_co_u32_e64 v1, s[6:7], v1, v3
; GFX9-O0-NEXT:    v_subb_co_u32_e64 v0, s[6:7], v0, v2, s[6:7]
; GFX9-O0-NEXT:    ; kill: def $vgpr1 killed $vgpr1 def $vgpr1_vgpr2 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v1
; GFX9-O0-NEXT:    v_lshrrev_b64 v[1:2], s4, v[1:2]
; GFX9-O0-NEXT:    ; kill: def $vgpr1 killed $vgpr1 killed $vgpr1_vgpr2 killed $exec
; GFX9-O0-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-O3-LABEL: called_i64:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O3-NEXT:    v_add_co_u32_e32 v2, vcc, v0, v0
; GFX9-O3-NEXT:    v_addc_co_u32_e32 v3, vcc, v1, v1, vcc
; GFX9-O3-NEXT:    v_mul_lo_u32 v4, v3, v0
; GFX9-O3-NEXT:    v_mul_lo_u32 v5, v2, v1
; GFX9-O3-NEXT:    v_mad_u64_u32 v[0:1], s[4:5], v2, v0, 0
; GFX9-O3-NEXT:    v_add3_u32 v1, v1, v5, v4
; GFX9-O3-NEXT:    v_sub_co_u32_e32 v0, vcc, v0, v2
; GFX9-O3-NEXT:    v_subb_co_u32_e32 v1, vcc, v1, v3, vcc
; GFX9-O3-NEXT:    s_setpc_b64 s[30:31]
  %add = add i64 %a, %a
  %mul = mul i64 %add, %a
  %sub = sub i64 %mul, %add
  ret i64 %sub
}

define amdgpu_kernel void @call_i64(ptr addrspace(8) inreg %tmp14, i64 inreg %arg) {
; GFX9-O0-LABEL: call_i64:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s32, 0x400
; GFX9-O0-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s26, -1
; GFX9-O0-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O0-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr12 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[10:11], -1
; GFX9-O0-NEXT:    v_writelane_b32 v12, s10, 0
; GFX9-O0-NEXT:    v_writelane_b32 v12, s11, 1
; GFX9-O0-NEXT:    s_mov_b32 s14, s8
; GFX9-O0-NEXT:    s_mov_b32 s13, s7
; GFX9-O0-NEXT:    s_mov_b32 s12, s6
; GFX9-O0-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[2:3]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v12, 0
; GFX9-O0-NEXT:    v_readlane_b32 s3, v12, 1
; GFX9-O0-NEXT:    v_writelane_b32 v12, s4, 2
; GFX9-O0-NEXT:    v_writelane_b32 v12, s5, 3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v12, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O0-NEXT:    v_readlane_b32 s0, v12, 2
; GFX9-O0-NEXT:    v_readlane_b32 s1, v12, 3
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x24
; GFX9-O0-NEXT:    s_load_dwordx2 s[18:19], s[0:1], 0x2c
; GFX9-O0-NEXT:    s_load_dwordx2 s[2:3], s[0:1], 0x34
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[8:9], -1
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s8, 4
; GFX9-O0-NEXT:    v_writelane_b32 v0, s9, 5
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_mov_b32 s6, s19
; GFX9-O0-NEXT:    s_mov_b32 s7, s18
; GFX9-O0-NEXT:    s_mov_b32 s15, s17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 killed $sgpr16_sgpr17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 def $sgpr16_sgpr17_sgpr18_sgpr19
; GFX9-O0-NEXT:    s_mov_b32 s17, s15
; GFX9-O0-NEXT:    s_mov_b32 s18, s7
; GFX9-O0-NEXT:    s_mov_b32 s19, s6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s16, 6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s17, 7
; GFX9-O0-NEXT:    v_writelane_b32 v0, s18, 8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s19, 9
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 0
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, s6
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, s7
; GFX9-O0-NEXT:    s_mov_b64 exec, s[8:9]
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, s2
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, s3
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 60
; GFX9-O0-NEXT:    s_mov_b32 s2, s0
; GFX9-O0-NEXT:    s_mov_b32 s0, s1
; GFX9-O0-NEXT:    s_mov_b32 s3, s6
; GFX9-O0-NEXT:    s_mov_b32 s1, s7
; GFX9-O0-NEXT:    s_add_u32 s8, s2, s3
; GFX9-O0-NEXT:    s_addc_u32 s0, s0, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s0
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v8
; GFX9-O0-NEXT:    s_mov_b32 s0, 32
; GFX9-O0-NEXT:    ; implicit-def: $sgpr2_sgpr3
; GFX9-O0-NEXT:    v_lshrrev_b64 v[10:11], s0, v[8:9]
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v10
; GFX9-O0-NEXT:    s_getpc_b64 s[0:1]
; GFX9-O0-NEXT:    s_add_u32 s0, s0, called_i64@gotpcrel32@lo+4
; GFX9-O0-NEXT:    s_addc_u32 s1, s1, called_i64@gotpcrel32@hi+12
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x0
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O0-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O0-NEXT:    s_mov_b32 s6, 20
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s6, v3
; GFX9-O0-NEXT:    s_mov_b32 s6, 10
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v4, s6, v4
; GFX9-O0-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6_sgpr7
; GFX9-O0-NEXT:    ; implicit-def: $sgpr15
; GFX9-O0-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_swappc_b64 s[30:31], s[16:17]
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v2, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s0, v2, 6
; GFX9-O0-NEXT:    v_readlane_b32 s1, v2, 7
; GFX9-O0-NEXT:    v_readlane_b32 s2, v2, 8
; GFX9-O0-NEXT:    v_readlane_b32 s3, v2, 9
; GFX9-O0-NEXT:    v_readlane_b32 s4, v2, 4
; GFX9-O0-NEXT:    v_readlane_b32 s5, v2, 5
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v8
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v9
; GFX9-O0-NEXT:    v_add_co_u32_e64 v3, s[6:7], v3, v5
; GFX9-O0-NEXT:    v_addc_co_u32_e64 v4, s[6:7], v4, v6, s[6:7]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v4
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_store_dwordx2 v[1:2], off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: call_i64:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O3-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O3-NEXT:    s_mov_b32 s26, -1
; GFX9-O3-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O3-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O3-NEXT:    s_mov_b32 s32, 0
; GFX9-O3-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O3-NEXT:    s_mov_b32 s14, s8
; GFX9-O3-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O3-NEXT:    s_load_dwordx2 s[4:5], s[2:3], 0x34
; GFX9-O3-NEXT:    s_load_dwordx4 s[16:19], s[2:3], 0x24
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, 0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s4
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, s5
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_add_u32 s8, s2, 60
; GFX9-O3-NEXT:    s_addc_u32 s9, s3, 0
; GFX9-O3-NEXT:    s_getpc_b64 s[2:3]
; GFX9-O3-NEXT:    s_add_u32 s2, s2, called_i64@gotpcrel32@lo+4
; GFX9-O3-NEXT:    s_addc_u32 s3, s3, called_i64@gotpcrel32@hi+12
; GFX9-O3-NEXT:    s_load_dwordx2 s[22:23], s[2:3], 0x0
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v3, 20, v3
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 10, v4
; GFX9-O3-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O3-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O3-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O3-NEXT:    s_mov_b32 s12, s6
; GFX9-O3-NEXT:    s_mov_b32 s13, s7
; GFX9-O3-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O3-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    s_swappc_b64 s[30:31], s[22:23]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_add_co_u32_e32 v3, vcc, v3, v6
; GFX9-O3-NEXT:    v_addc_co_u32_e32 v4, vcc, v4, v7, vcc
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v4
; GFX9-O3-NEXT:    buffer_store_dwordx2 v[0:1], off, s[16:19], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm


  %tmp107 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %arg, i64 0)
  %tmp134 = call i64 @called_i64(i64 %tmp107)
  %tmp136 = add i64 %tmp134, %tmp107
  %tmp137 = tail call i64 @llvm.amdgcn.wwm.i64(i64 %tmp136)
  %tmp138 = bitcast i64 %tmp137 to <2 x i32>
  call void @llvm.amdgcn.raw.ptr.buffer.store.v2i32(<2 x i32> %tmp138, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define amdgpu_cs void @_amdgpu_cs_main(<4 x i32> inreg %desc, i32 %index) {
; GFX9-O0-LABEL: _amdgpu_cs_main:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s4, s3
; GFX9-O0-NEXT:    s_mov_b32 s5, s2
; GFX9-O0-NEXT:    s_mov_b32 s6, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s6
; GFX9-O0-NEXT:    s_mov_b32 s2, s5
; GFX9-O0-NEXT:    s_mov_b32 s3, s4
; GFX9-O0-NEXT:    ; kill: def $sgpr4_sgpr5_sgpr6_sgpr7 killed $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s4, 5
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v0, s4, v0
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_load_dwordx4 v[10:13], v0, s[0:3], s4 offen
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[3:4], v0, s[0:3], s4 offen offset:16
; GFX9-O0-NEXT:    s_waitcnt vmcnt(1)
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v11
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v10
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v7
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    s_mov_b32 s5, 0x7fffffff
; GFX9-O0-NEXT:    s_mov_b32 s8, -1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s5
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v13
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v12
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v7
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v2
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v11, v9
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v8
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, v7
; GFX9-O0-NEXT:    v_mov_b32_e32 v10, v6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6_vgpr7_vgpr8 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v11
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v10
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, v9
; GFX9-O0-NEXT:    buffer_store_dwordx4 v[5:8], v0, s[0:3], s4 offen
; GFX9-O0-NEXT:    buffer_store_dwordx2 v[3:4], v0, s[0:3], s4 offen offset:16
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: _amdgpu_cs_main:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v0, 5, v0
; GFX9-O3-NEXT:    buffer_load_dwordx4 v[7:10], v0, s[0:3], 0 offen
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[11:12], v0, s[0:3], 0 offen offset:16
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    s_mov_b32 s6, -1
; GFX9-O3-NEXT:    s_brev_b32 s7, -2
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v2, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    s_waitcnt vmcnt(1)
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O3-NEXT:    v_mov_b32_e32 v2, v8
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v9
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v10
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v11
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, v12
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v9, v3
; GFX9-O3-NEXT:    v_mov_b32_e32 v8, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v10, v4
; GFX9-O3-NEXT:    v_mov_b32_e32 v11, v5
; GFX9-O3-NEXT:    v_mov_b32_e32 v12, v6
; GFX9-O3-NEXT:    buffer_store_dwordx4 v[7:10], v0, s[0:3], 0 offen
; GFX9-O3-NEXT:    buffer_store_dwordx2 v[11:12], v0, s[0:3], 0 offen offset:16
; GFX9-O3-NEXT:    s_endpgm
  %tmp17 = shl i32 %index, 5
  %tmp18 = tail call <4 x i32> @llvm.amdgcn.s.buffer.load.v4i32(<4 x i32> %desc, i32 %tmp17, i32 0)
  %.i0.upto1.bc = bitcast <4 x i32> %tmp18 to <2 x i64>
  %tmp19 = or i32 %tmp17, 16
  %tmp20 = tail call <2 x i32> @llvm.amdgcn.s.buffer.load.v2i32(<4 x i32> %desc, i32 %tmp19, i32 0)
  %.i0.upto1.extract = extractelement <2 x i64> %.i0.upto1.bc, i32 0
  %tmp22 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i0.upto1.extract, i64 9223372036854775807)
  %tmp97 = tail call i64 @llvm.amdgcn.wwm.i64(i64 %tmp22)
  %.i1.upto1.extract = extractelement <2 x i64> %.i0.upto1.bc, i32 1
  %tmp99 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i1.upto1.extract, i64 9223372036854775807)
  %tmp174 = tail call i64 @llvm.amdgcn.wwm.i64(i64 %tmp99)
  %.i25 = bitcast <2 x i32> %tmp20 to i64
  %tmp176 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i25, i64 9223372036854775807)
  %tmp251 = tail call i64 @llvm.amdgcn.wwm.i64(i64 %tmp176)
  %.cast = bitcast i64 %tmp97 to <2 x float>
  %.cast6 = bitcast i64 %tmp174 to <2 x float>
  %.cast7 = bitcast i64 %tmp251 to <2 x float>
  %tmp254 = shufflevector <2 x float> %.cast, <2 x float> %.cast6, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %desc.int = bitcast <4 x i32> %desc to i128
  %desc.ptr = inttoptr i128 %desc.int to ptr addrspace(8)
  tail call void @llvm.amdgcn.raw.ptr.buffer.store.v4f32(<4 x float> %tmp254, ptr addrspace(8) %desc.ptr, i32 %tmp17, i32 0, i32 0)
  tail call void @llvm.amdgcn.raw.ptr.buffer.store.v2f32(<2 x float> %.cast7, ptr addrspace(8) %desc.ptr, i32 %tmp19, i32 0, i32 0)
  ret void
}


define amdgpu_cs void @strict_wwm_no_cfg(ptr addrspace(8) inreg %tmp14) {
; GFX9-O0-LABEL: strict_wwm_no_cfg:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s6, s2
; GFX9-O0-NEXT:    s_mov_b32 s4, s0
; GFX9-O0-NEXT:    ; kill: def $sgpr6 killed $sgpr6 def $sgpr6_sgpr7
; GFX9-O0-NEXT:    s_mov_b32 s7, s3
; GFX9-O0-NEXT:    s_mov_b32 s8, s7
; GFX9-O0-NEXT:    s_mov_b32 s9, s6
; GFX9-O0-NEXT:    ; kill: def $sgpr4 killed $sgpr4 def $sgpr4_sgpr5
; GFX9-O0-NEXT:    s_mov_b32 s5, s1
; GFX9-O0-NEXT:    s_mov_b32 s10, s5
; GFX9-O0-NEXT:    s_mov_b32 s0, s4
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s10
; GFX9-O0-NEXT:    s_mov_b32 s2, s9
; GFX9-O0-NEXT:    s_mov_b32 s3, s8
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[5:6], off, s[0:3], s4
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v6
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, s4
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s4
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s4
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v2, v0 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v0, v0, v2
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, s4
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v0, v1 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v0, v1, v0
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v0
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[6:7], v3, v4
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v3, 0, 1, s[6:7]
; GFX9-O0-NEXT:    s_mov_b32 s5, 1
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s5, v3
; GFX9-O0-NEXT:    s_mov_b32 s5, 2
; GFX9-O0-NEXT:    v_and_b32_e64 v3, v3, s5
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: strict_wwm_no_cfg:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[4:5], off, s[0:3], 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v4, s[4:5]
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v3, 0, v5, s[4:5]
; GFX9-O3-NEXT:    s_nop 0
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_mov_b32_dpp v0, v3 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    v_add_u32_e32 v0, v3, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, v4, v5
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v4, 0, 1, vcc
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 1, v4
; GFX9-O3-NEXT:    v_and_b32_e32 v4, 2, v4
; GFX9-O3-NEXT:    buffer_store_dword v4, off, s[0:3], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm
  %tmp100 = call <2 x float> @llvm.amdgcn.raw.ptr.buffer.load.v2f32(ptr addrspace(8) %tmp14, i32 0, i32 0, i32 0)
  %tmp101 = bitcast <2 x float> %tmp100 to <2 x i32>
  %tmp102 = extractelement <2 x i32> %tmp101, i32 0
  %tmp103 = extractelement <2 x i32> %tmp101, i32 1
  %tmp105 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp102, i32 0)
  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp103, i32 0)


  %tmp120 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp105, i32 323, i32 12, i32 15, i1 false)
  %tmp121 = add i32 %tmp105, %tmp120
  %tmp122 = tail call i32 @llvm.amdgcn.strict.wwm.i32(i32 %tmp121)

  %tmp135 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp107, i32 323, i32 12, i32 15, i1 false)
  %tmp136 = add i32 %tmp107, %tmp135
  %tmp137 = tail call i32 @llvm.amdgcn.strict.wwm.i32(i32 %tmp136)

  %tmp138 = icmp eq i32 %tmp122, %tmp137
  %tmp139 = sext i1 %tmp138 to i32
  %tmp140 = shl nsw i32 %tmp139, 1
  %tmp141 = and i32 %tmp140, 2
  %tmp145 = bitcast i32 %tmp141 to float
  call void @llvm.amdgcn.raw.ptr.buffer.store.f32(float %tmp145, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define amdgpu_cs void @strict_wwm_cfg(ptr addrspace(8) inreg %tmp14, i32 %arg) {
; GFX9-O0-LABEL: strict_wwm_cfg:
; GFX9-O0:       ; %bb.0: ; %entry
; GFX9-O0-NEXT:    s_mov_b32 s16, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s17, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s18, -1
; GFX9-O0-NEXT:    s_mov_b32 s19, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s16, s16, s4
; GFX9-O0-NEXT:    s_addc_u32 s17, s17, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr3 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[16:19], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 0
; GFX9-O0-NEXT:    s_mov_b32 s4, s1
; GFX9-O0-NEXT:    v_readlane_b32 s1, v0, 0
; GFX9-O0-NEXT:    ; kill: def $sgpr2 killed $sgpr2 def $sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s3, s1
; GFX9-O0-NEXT:    s_mov_b32 s8, s3
; GFX9-O0-NEXT:    s_mov_b32 s9, s2
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1
; GFX9-O0-NEXT:    s_mov_b32 s1, s4
; GFX9-O0-NEXT:    s_mov_b32 s10, s1
; GFX9-O0-NEXT:    s_mov_b32 s4, s0
; GFX9-O0-NEXT:    ; kill: def $sgpr4 killed $sgpr4 def $sgpr4_sgpr5_sgpr6_sgpr7
; GFX9-O0-NEXT:    s_mov_b32 s5, s10
; GFX9-O0-NEXT:    s_mov_b32 s6, s9
; GFX9-O0-NEXT:    s_mov_b32 s7, s8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s2, 1
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 2
; GFX9-O0-NEXT:    v_writelane_b32 v0, s0, 3
; GFX9-O0-NEXT:    v_writelane_b32 v0, s1, 4
; GFX9-O0-NEXT:    s_mov_b32 s0, 0
; GFX9-O0-NEXT:    s_nop 2
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[4:5], off, s[4:7], s0
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    buffer_store_dword v4, off, s[16:19], 0 offset:12 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    buffer_store_dword v5, off, s[16:19], 0 offset:16 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    ; implicit-def: $sgpr2_sgpr3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[2:3], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s0
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s0
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v2, v1 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v1, v2
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    buffer_store_dword v4, off, s[16:19], 0 offset:8 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[2:3], v3, s0
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, s0
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[16:19], 0 offset:4 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], exec
; GFX9-O0-NEXT:    v_writelane_b32 v0, s0, 5
; GFX9-O0-NEXT:    v_writelane_b32 v0, s1, 6
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[16:19], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_and_b64 s[0:1], s[0:1], s[2:3]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[0:1]
; GFX9-O0-NEXT:    s_cbranch_execz .LBB8_2
; GFX9-O0-NEXT:  ; %bb.1: ; %if
; GFX9-O0-NEXT:    buffer_load_dword v3, off, s[16:19], 0 offset:12 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    buffer_load_dword v4, off, s[16:19], 0 offset:16 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[0:1], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v2, v1, v0, s[0:1]
; GFX9-O0-NEXT:    s_nop 1
; GFX9-O0-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v2, v1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[16:19], 0 offset:4 ; 4-byte Folded Spill
; GFX9-O0-NEXT:  .LBB8_2: ; %merge
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[16:19], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s4, v0, 5
; GFX9-O0-NEXT:    v_readlane_b32 s5, v0, 6
; GFX9-O0-NEXT:    s_or_b64 exec, exec, s[4:5]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v0, 1
; GFX9-O0-NEXT:    v_readlane_b32 s3, v0, 2
; GFX9-O0-NEXT:    v_readlane_b32 s0, v0, 3
; GFX9-O0-NEXT:    v_readlane_b32 s1, v0, 4
; GFX9-O0-NEXT:    buffer_load_dword v3, off, s[16:19], 0 offset:8 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    buffer_load_dword v4, off, s[16:19], 0 offset:4 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_cmp_eq_u32_e64 s[4:5], v3, v4
; GFX9-O0-NEXT:    v_cndmask_b32_e64 v3, 0, 1, s[4:5]
; GFX9-O0-NEXT:    s_mov_b32 s4, 1
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s4, v3
; GFX9-O0-NEXT:    s_mov_b32 s4, 2
; GFX9-O0-NEXT:    v_and_b32_e64 v3, v3, s4
; GFX9-O0-NEXT:    s_mov_b32 s6, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 killed $sgpr0_sgpr1
; GFX9-O0-NEXT:    s_mov_b32 s4, s3
; GFX9-O0-NEXT:    s_mov_b32 s5, s2
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s6
; GFX9-O0-NEXT:    s_mov_b32 s2, s5
; GFX9-O0-NEXT:    s_mov_b32 s3, s4
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_store_dword v3, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: strict_wwm_cfg:
; GFX9-O3:       ; %bb.0: ; %entry
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[3:4], off, s[0:3], 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v3, s[4:5]
; GFX9-O3-NEXT:    s_nop 1
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v0
; GFX9-O3-NEXT:    s_and_saveexec_b64 s[4:5], vcc
; GFX9-O3-NEXT:  ; %bb.1: ; %if
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v2, 0, v4, s[6:7]
; GFX9-O3-NEXT:    s_nop 1
; GFX9-O3-NEXT:    v_mov_b32_dpp v1, v2 row_bcast:31 row_mask:0xc bank_mask:0xf
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v2, v1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v1
; GFX9-O3-NEXT:  ; %bb.2: ; %merge
; GFX9-O3-NEXT:    s_or_b64 exec, exec, s[4:5]
; GFX9-O3-NEXT:    v_cmp_eq_u32_e32 vcc, v3, v5
; GFX9-O3-NEXT:    v_cndmask_b32_e64 v0, 0, 1, vcc
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v0, 1, v0
; GFX9-O3-NEXT:    v_and_b32_e32 v0, 2, v0
; GFX9-O3-NEXT:    buffer_store_dword v0, off, s[0:3], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm
entry:
  %tmp100 = call <2 x float> @llvm.amdgcn.raw.ptr.buffer.load.v2f32(ptr addrspace(8) %tmp14, i32 0, i32 0, i32 0)
  %tmp101 = bitcast <2 x float> %tmp100 to <2 x i32>
  %tmp102 = extractelement <2 x i32> %tmp101, i32 0
  %tmp105 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp102, i32 0)

  %tmp120 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp105, i32 323, i32 12, i32 15, i1 false)
  %tmp121 = add i32 %tmp105, %tmp120
  %tmp122 = tail call i32 @llvm.amdgcn.strict.wwm.i32(i32 %tmp121)

  %cond = icmp eq i32 %arg, 0
  br i1 %cond, label %if, label %merge
if:
  %tmp103 = extractelement <2 x i32> %tmp101, i32 1
  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %tmp103, i32 0)

  %tmp135 = tail call i32 @llvm.amdgcn.update.dpp.i32(i32 0, i32 %tmp107, i32 323, i32 12, i32 15, i1 false)
  %tmp136 = add i32 %tmp107, %tmp135
  %tmp137 = tail call i32 @llvm.amdgcn.strict.wwm.i32(i32 %tmp136)
  br label %merge

merge:
  %merge_value = phi i32 [ 0, %entry ], [%tmp137, %if ]
  %tmp138 = icmp eq i32 %tmp122, %merge_value
  %tmp139 = sext i1 %tmp138 to i32
  %tmp140 = shl nsw i32 %tmp139, 1
  %tmp141 = and i32 %tmp140, 2
  %tmp145 = bitcast i32 %tmp141 to float
  call void @llvm.amdgcn.raw.ptr.buffer.store.f32(float %tmp145, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define hidden i32 @strict_wwm_called(i32 %a) noinline {
; GFX9-O0-LABEL: strict_wwm_called:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O0-NEXT:    v_add_u32_e64 v1, v0, v0
; GFX9-O0-NEXT:    v_mul_lo_u32 v0, v1, v0
; GFX9-O0-NEXT:    v_sub_u32_e64 v0, v0, v1
; GFX9-O0-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-O3-LABEL: strict_wwm_called:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O3-NEXT:    v_add_u32_e32 v1, v0, v0
; GFX9-O3-NEXT:    v_mul_lo_u32 v0, v1, v0
; GFX9-O3-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-O3-NEXT:    s_setpc_b64 s[30:31]
  %add = add i32 %a, %a
  %mul = mul i32 %add, %a
  %sub = sub i32 %mul, %add
  ret i32 %sub
}

define amdgpu_kernel void @strict_wwm_call(ptr addrspace(8) inreg %tmp14, i32 inreg %arg) {
; GFX9-O0-LABEL: strict_wwm_call:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s32, 0x400
; GFX9-O0-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s26, -1
; GFX9-O0-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O0-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr7 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[10:11], -1
; GFX9-O0-NEXT:    v_writelane_b32 v7, s10, 0
; GFX9-O0-NEXT:    v_writelane_b32 v7, s11, 1
; GFX9-O0-NEXT:    s_mov_b32 s14, s8
; GFX9-O0-NEXT:    s_mov_b32 s13, s7
; GFX9-O0-NEXT:    s_mov_b32 s12, s6
; GFX9-O0-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[2:3]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v7, 0
; GFX9-O0-NEXT:    v_readlane_b32 s3, v7, 1
; GFX9-O0-NEXT:    v_writelane_b32 v7, s4, 2
; GFX9-O0-NEXT:    v_writelane_b32 v7, s5, 3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v7, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O0-NEXT:    v_readlane_b32 s0, v7, 2
; GFX9-O0-NEXT:    v_readlane_b32 s1, v7, 3
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x24
; GFX9-O0-NEXT:    s_load_dwordx2 s[8:9], s[0:1], 0x2c
; GFX9-O0-NEXT:    s_load_dword s2, s[0:1], 0x34
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s6, 4
; GFX9-O0-NEXT:    v_writelane_b32 v0, s7, 5
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_mov_b32 s3, s9
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 killed $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 killed $sgpr16_sgpr17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 def $sgpr16_sgpr17_sgpr18_sgpr19
; GFX9-O0-NEXT:    s_mov_b32 s17, s9
; GFX9-O0-NEXT:    s_mov_b32 s18, s8
; GFX9-O0-NEXT:    s_mov_b32 s19, s3
; GFX9-O0-NEXT:    v_writelane_b32 v0, s16, 6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s17, 7
; GFX9-O0-NEXT:    v_writelane_b32 v0, s18, 8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s19, 9
; GFX9-O0-NEXT:    s_mov_b32 s3, 0
; GFX9-O0-NEXT:    v_writelane_b32 v0, s3, 10
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, s3
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, s2
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 56
; GFX9-O0-NEXT:    s_mov_b32 s2, s0
; GFX9-O0-NEXT:    s_mov_b32 s0, s1
; GFX9-O0-NEXT:    s_mov_b32 s3, s6
; GFX9-O0-NEXT:    s_mov_b32 s1, s7
; GFX9-O0-NEXT:    s_add_u32 s8, s2, s3
; GFX9-O0-NEXT:    s_addc_u32 s0, s0, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s0
; GFX9-O0-NEXT:    s_getpc_b64 s[16:17]
; GFX9-O0-NEXT:    s_add_u32 s16, s16, strict_wwm_called@rel32@lo+4
; GFX9-O0-NEXT:    s_addc_u32 s17, s17, strict_wwm_called@rel32@hi+12
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O0-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O0-NEXT:    s_mov_b32 s6, 20
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s6, v3
; GFX9-O0-NEXT:    s_mov_b32 s6, 10
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v4, s6, v4
; GFX9-O0-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6_sgpr7
; GFX9-O0-NEXT:    ; implicit-def: $sgpr15
; GFX9-O0-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O0-NEXT:    s_swappc_b64 s[30:31], s[16:17]
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v1, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s0, v1, 6
; GFX9-O0-NEXT:    v_readlane_b32 s1, v1, 7
; GFX9-O0-NEXT:    v_readlane_b32 s2, v1, 8
; GFX9-O0-NEXT:    v_readlane_b32 s3, v1, 9
; GFX9-O0-NEXT:    v_readlane_b32 s6, v1, 4
; GFX9-O0-NEXT:    v_readlane_b32 s7, v1, 5
; GFX9-O0-NEXT:    v_readlane_b32 s4, v1, 10
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_add_u32_e64 v3, v3, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    buffer_store_dword v1, off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: strict_wwm_call:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O3-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O3-NEXT:    s_mov_b32 s26, -1
; GFX9-O3-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O3-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O3-NEXT:    s_mov_b32 s32, 0
; GFX9-O3-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O3-NEXT:    s_mov_b32 s14, s8
; GFX9-O3-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O3-NEXT:    s_load_dword s4, s[2:3], 0x34
; GFX9-O3-NEXT:    s_load_dwordx4 s[16:19], s[2:3], 0x24
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, 0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s4
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_add_u32 s8, s2, 56
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v3, 20, v3
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 10, v4
; GFX9-O3-NEXT:    s_addc_u32 s9, s3, 0
; GFX9-O3-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O3-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O3-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O3-NEXT:    s_mov_b32 s12, s6
; GFX9-O3-NEXT:    s_mov_b32 s13, s7
; GFX9-O3-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O3-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O3-NEXT:    s_getpc_b64 s[22:23]
; GFX9-O3-NEXT:    s_add_u32 s22, s22, strict_wwm_called@rel32@lo+4
; GFX9-O3-NEXT:    s_addc_u32 s23, s23, strict_wwm_called@rel32@hi+12
; GFX9-O3-NEXT:    s_swappc_b64 s[30:31], s[22:23]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O3-NEXT:    v_add_u32_e32 v3, v3, v6
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O3-NEXT:    buffer_store_dword v0, off, s[16:19], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm


  %tmp107 = tail call i32 @llvm.amdgcn.set.inactive.i32(i32 %arg, i32 0)
  %tmp134 = call i32 @strict_wwm_called(i32 %tmp107)
  %tmp136 = add i32 %tmp134, %tmp107
  %tmp137 = tail call i32 @llvm.amdgcn.strict.wwm.i32(i32 %tmp136)
  call void @llvm.amdgcn.raw.ptr.buffer.store.i32(i32 %tmp137, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define i64 @strict_wwm_called_i64(i64 %a) noinline {
; GFX9-O0-LABEL: strict_wwm_called_i64:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v0
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4
; GFX9-O0-NEXT:    ; kill: def $vgpr2 killed $vgpr2 def $vgpr2_vgpr3 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O0-NEXT:    ; implicit-def: $sgpr4_sgpr5
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_add_co_u32_e64 v4, s[4:5], v4, v5
; GFX9-O0-NEXT:    v_addc_co_u32_e64 v0, s[4:5], v0, v1, s[4:5]
; GFX9-O0-NEXT:    ; kill: def $vgpr4 killed $vgpr4 def $vgpr4_vgpr5 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_mov_b32 s4, 32
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_lshrrev_b64 v[0:1], s4, v[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v4
; GFX9-O0-NEXT:    v_mul_lo_u32 v1, v0, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_lshrrev_b64 v[6:7], s4, v[4:5]
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    v_mul_lo_u32 v2, v2, v3
; GFX9-O0-NEXT:    v_mad_u64_u32 v[6:7], s[6:7], v0, v3, 0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_add3_u32 v0, v0, v1, v2
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s5
; GFX9-O0-NEXT:    ; kill: def $vgpr0 killed $vgpr0 def $vgpr0_vgpr1 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v2
; GFX9-O0-NEXT:    v_lshlrev_b64 v[1:2], s4, v[0:1]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 killed $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    s_mov_b32 s5, 0
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, 0
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 def $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_or_b32_e64 v0, v0, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v6
; GFX9-O0-NEXT:    v_or_b32_e64 v6, v1, v2
; GFX9-O0-NEXT:    ; kill: def $vgpr6 killed $vgpr6 def $vgpr6_vgpr7 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v6
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v4
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v7
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v5
; GFX9-O0-NEXT:    v_sub_co_u32_e64 v1, s[6:7], v1, v3
; GFX9-O0-NEXT:    v_subb_co_u32_e64 v0, s[6:7], v0, v2, s[6:7]
; GFX9-O0-NEXT:    ; kill: def $vgpr1 killed $vgpr1 def $vgpr1_vgpr2 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v0
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v1
; GFX9-O0-NEXT:    v_lshrrev_b64 v[1:2], s4, v[1:2]
; GFX9-O0-NEXT:    ; kill: def $vgpr1 killed $vgpr1 killed $vgpr1_vgpr2 killed $exec
; GFX9-O0-NEXT:    s_setpc_b64 s[30:31]
;
; GFX9-O3-LABEL: strict_wwm_called_i64:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
; GFX9-O3-NEXT:    v_add_co_u32_e32 v2, vcc, v0, v0
; GFX9-O3-NEXT:    v_addc_co_u32_e32 v3, vcc, v1, v1, vcc
; GFX9-O3-NEXT:    v_mul_lo_u32 v4, v3, v0
; GFX9-O3-NEXT:    v_mul_lo_u32 v5, v2, v1
; GFX9-O3-NEXT:    v_mad_u64_u32 v[0:1], s[4:5], v2, v0, 0
; GFX9-O3-NEXT:    v_add3_u32 v1, v1, v5, v4
; GFX9-O3-NEXT:    v_sub_co_u32_e32 v0, vcc, v0, v2
; GFX9-O3-NEXT:    v_subb_co_u32_e32 v1, vcc, v1, v3, vcc
; GFX9-O3-NEXT:    s_setpc_b64 s[30:31]
  %add = add i64 %a, %a
  %mul = mul i64 %add, %a
  %sub = sub i64 %mul, %add
  ret i64 %sub
}

define amdgpu_kernel void @strict_wwm_call_i64(ptr addrspace(8) inreg %tmp14, i64 inreg %arg) {
; GFX9-O0-LABEL: strict_wwm_call_i64:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s32, 0x400
; GFX9-O0-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O0-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O0-NEXT:    s_mov_b32 s26, -1
; GFX9-O0-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O0-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O0-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O0-NEXT:    ; implicit-def: $vgpr12 : SGPR spill to VGPR lane
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[10:11], -1
; GFX9-O0-NEXT:    v_writelane_b32 v12, s10, 0
; GFX9-O0-NEXT:    v_writelane_b32 v12, s11, 1
; GFX9-O0-NEXT:    s_mov_b32 s14, s8
; GFX9-O0-NEXT:    s_mov_b32 s13, s7
; GFX9-O0-NEXT:    s_mov_b32 s12, s6
; GFX9-O0-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[2:3]
; GFX9-O0-NEXT:    v_readlane_b32 s2, v12, 0
; GFX9-O0-NEXT:    v_readlane_b32 s3, v12, 1
; GFX9-O0-NEXT:    v_writelane_b32 v12, s4, 2
; GFX9-O0-NEXT:    v_writelane_b32 v12, s5, 3
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v12, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O0-NEXT:    v_readlane_b32 s0, v12, 2
; GFX9-O0-NEXT:    v_readlane_b32 s1, v12, 3
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[2:3]
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x24
; GFX9-O0-NEXT:    s_load_dwordx2 s[18:19], s[0:1], 0x2c
; GFX9-O0-NEXT:    s_load_dwordx2 s[2:3], s[0:1], 0x34
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[8:9], -1
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_writelane_b32 v0, s8, 4
; GFX9-O0-NEXT:    v_writelane_b32 v0, s9, 5
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_mov_b32 s6, s19
; GFX9-O0-NEXT:    s_mov_b32 s7, s18
; GFX9-O0-NEXT:    s_mov_b32 s15, s17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 killed $sgpr16_sgpr17
; GFX9-O0-NEXT:    ; kill: def $sgpr16 killed $sgpr16 def $sgpr16_sgpr17_sgpr18_sgpr19
; GFX9-O0-NEXT:    s_mov_b32 s17, s15
; GFX9-O0-NEXT:    s_mov_b32 s18, s7
; GFX9-O0-NEXT:    s_mov_b32 s19, s6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s16, 6
; GFX9-O0-NEXT:    v_writelane_b32 v0, s17, 7
; GFX9-O0-NEXT:    v_writelane_b32 v0, s18, 8
; GFX9-O0-NEXT:    v_writelane_b32 v0, s19, 9
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_store_dword v0, off, s[24:27], 0 ; 4-byte Folded Spill
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 0
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, s6
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, s7
; GFX9-O0-NEXT:    s_mov_b64 exec, s[8:9]
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, s2
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, s3
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 s[6:7], 60
; GFX9-O0-NEXT:    s_mov_b32 s2, s0
; GFX9-O0-NEXT:    s_mov_b32 s0, s1
; GFX9-O0-NEXT:    s_mov_b32 s3, s6
; GFX9-O0-NEXT:    s_mov_b32 s1, s7
; GFX9-O0-NEXT:    s_add_u32 s8, s2, s3
; GFX9-O0-NEXT:    s_addc_u32 s0, s0, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s0
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v8
; GFX9-O0-NEXT:    s_mov_b32 s0, 32
; GFX9-O0-NEXT:    ; implicit-def: $sgpr2_sgpr3
; GFX9-O0-NEXT:    v_lshrrev_b64 v[10:11], s0, v[8:9]
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v10
; GFX9-O0-NEXT:    s_getpc_b64 s[0:1]
; GFX9-O0-NEXT:    s_add_u32 s0, s0, strict_wwm_called_i64@gotpcrel32@lo+4
; GFX9-O0-NEXT:    s_addc_u32 s1, s1, strict_wwm_called_i64@gotpcrel32@hi+12
; GFX9-O0-NEXT:    s_load_dwordx2 s[16:17], s[0:1], 0x0
; GFX9-O0-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O0-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O0-NEXT:    s_mov_b32 s6, 20
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v3, s6, v3
; GFX9-O0-NEXT:    s_mov_b32 s6, 10
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v4, s6, v4
; GFX9-O0-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6_sgpr7
; GFX9-O0-NEXT:    ; implicit-def: $sgpr15
; GFX9-O0-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O0-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O0-NEXT:    s_swappc_b64 s[30:31], s[16:17]
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v2, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_readlane_b32 s0, v2, 6
; GFX9-O0-NEXT:    v_readlane_b32 s1, v2, 7
; GFX9-O0-NEXT:    v_readlane_b32 s2, v2, 8
; GFX9-O0-NEXT:    v_readlane_b32 s3, v2, 9
; GFX9-O0-NEXT:    v_readlane_b32 s4, v2, 4
; GFX9-O0-NEXT:    v_readlane_b32 s5, v2, 5
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O0-NEXT:    buffer_load_dword v0, off, s[24:27], 0 ; 4-byte Folded Reload
; GFX9-O0-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr6
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v8
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v9
; GFX9-O0-NEXT:    v_add_co_u32_e64 v3, s[6:7], v3, v5
; GFX9-O0-NEXT:    v_addc_co_u32_e64 v4, s[6:7], v4, v6, s[6:7]
; GFX9-O0-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v4
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_store_dwordx2 v[1:2], off, s[0:3], s4 offset:4
; GFX9-O0-NEXT:    ; kill: killed $vgpr0
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: strict_wwm_call_i64:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    s_mov_b32 s24, SCRATCH_RSRC_DWORD0
; GFX9-O3-NEXT:    s_mov_b32 s25, SCRATCH_RSRC_DWORD1
; GFX9-O3-NEXT:    s_mov_b32 s26, -1
; GFX9-O3-NEXT:    s_mov_b32 s27, 0xe00000
; GFX9-O3-NEXT:    s_add_u32 s24, s24, s9
; GFX9-O3-NEXT:    s_mov_b32 s32, 0
; GFX9-O3-NEXT:    s_addc_u32 s25, s25, 0
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[12:13], -1
; GFX9-O3-NEXT:    s_mov_b32 s14, s8
; GFX9-O3-NEXT:    s_mov_b64 s[10:11], s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[12:13]
; GFX9-O3-NEXT:    s_load_dwordx2 s[4:5], s[2:3], 0x34
; GFX9-O3-NEXT:    s_load_dwordx4 s[16:19], s[2:3], 0x24
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[20:21], -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, 0
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, 0
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s4
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, s5
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_add_u32 s8, s2, 60
; GFX9-O3-NEXT:    s_addc_u32 s9, s3, 0
; GFX9-O3-NEXT:    s_getpc_b64 s[2:3]
; GFX9-O3-NEXT:    s_add_u32 s2, s2, strict_wwm_called_i64@gotpcrel32@lo+4
; GFX9-O3-NEXT:    s_addc_u32 s3, s3, strict_wwm_called_i64@gotpcrel32@hi+12
; GFX9-O3-NEXT:    s_load_dwordx2 s[22:23], s[2:3], 0x0
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v3, 20, v3
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v4, 10, v4
; GFX9-O3-NEXT:    v_or3_b32 v3, v5, v4, v3
; GFX9-O3-NEXT:    s_mov_b64 s[4:5], s[0:1]
; GFX9-O3-NEXT:    s_mov_b64 s[0:1], s[24:25]
; GFX9-O3-NEXT:    s_mov_b32 s12, s6
; GFX9-O3-NEXT:    s_mov_b32 s13, s7
; GFX9-O3-NEXT:    v_mov_b32_e32 v31, v3
; GFX9-O3-NEXT:    s_mov_b64 s[2:3], s[26:27]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v6
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O3-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-O3-NEXT:    s_swappc_b64 s[30:31], s[22:23]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v0
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v1
; GFX9-O3-NEXT:    v_add_co_u32_e32 v3, vcc, v3, v6
; GFX9-O3-NEXT:    v_addc_co_u32_e32 v4, vcc, v4, v7, vcc
; GFX9-O3-NEXT:    s_mov_b64 exec, s[20:21]
; GFX9-O3-NEXT:    v_mov_b32_e32 v0, v3
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v4
; GFX9-O3-NEXT:    buffer_store_dwordx2 v[0:1], off, s[16:19], 0 offset:4
; GFX9-O3-NEXT:    s_endpgm


  %tmp107 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %arg, i64 0)
  %tmp134 = call i64 @strict_wwm_called_i64(i64 %tmp107)
  %tmp136 = add i64 %tmp134, %tmp107
  %tmp137 = tail call i64 @llvm.amdgcn.strict.wwm.i64(i64 %tmp136)
  %tmp138 = bitcast i64 %tmp137 to <2 x i32>
  call void @llvm.amdgcn.raw.ptr.buffer.store.v2i32(<2 x i32> %tmp138, ptr addrspace(8) %tmp14, i32 4, i32 0, i32 0)
  ret void
}

define amdgpu_cs void @strict_wwm_amdgpu_cs_main(<4 x i32> inreg %desc, i32 %index) {
; GFX9-O0-LABEL: strict_wwm_amdgpu_cs_main:
; GFX9-O0:       ; %bb.0:
; GFX9-O0-NEXT:    s_mov_b32 s4, s3
; GFX9-O0-NEXT:    s_mov_b32 s5, s2
; GFX9-O0-NEXT:    s_mov_b32 s6, s1
; GFX9-O0-NEXT:    ; kill: def $sgpr0 killed $sgpr0 def $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s1, s6
; GFX9-O0-NEXT:    s_mov_b32 s2, s5
; GFX9-O0-NEXT:    s_mov_b32 s3, s4
; GFX9-O0-NEXT:    ; kill: def $sgpr4_sgpr5_sgpr6_sgpr7 killed $sgpr0_sgpr1_sgpr2_sgpr3
; GFX9-O0-NEXT:    s_mov_b32 s4, 5
; GFX9-O0-NEXT:    v_lshlrev_b32_e64 v0, s4, v0
; GFX9-O0-NEXT:    s_mov_b32 s4, 0
; GFX9-O0-NEXT:    buffer_load_dwordx4 v[10:13], v0, s[0:3], s4 offen
; GFX9-O0-NEXT:    buffer_load_dwordx2 v[3:4], v0, s[0:3], s4 offen offset:16
; GFX9-O0-NEXT:    s_waitcnt vmcnt(1)
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v11
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v10
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v7
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    s_mov_b32 s5, 0x7fffffff
; GFX9-O0-NEXT:    s_mov_b32 s8, -1
; GFX9-O0-NEXT:    ; kill: def $sgpr8 killed $sgpr8 def $sgpr8_sgpr9
; GFX9-O0-NEXT:    s_mov_b32 s9, s5
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v13
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v12
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v7
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v5
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v6
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v2
; GFX9-O0-NEXT:    s_or_saveexec_b64 s[6:7], -1
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, s8
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, s9
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O0-NEXT:    v_mov_b32_e32 v1, v3
; GFX9-O0-NEXT:    v_mov_b32_e32 v2, v4
; GFX9-O0-NEXT:    s_mov_b64 exec, -1
; GFX9-O0-NEXT:    s_mov_b64 exec, s[6:7]
; GFX9-O0-NEXT:    v_mov_b32_e32 v3, v1
; GFX9-O0-NEXT:    v_mov_b32_e32 v4, v2
; GFX9-O0-NEXT:    v_mov_b32_e32 v11, v9
; GFX9-O0-NEXT:    v_mov_b32_e32 v5, v8
; GFX9-O0-NEXT:    v_mov_b32_e32 v9, v7
; GFX9-O0-NEXT:    v_mov_b32_e32 v10, v6
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; implicit-def: $sgpr5
; GFX9-O0-NEXT:    ; kill: def $vgpr5 killed $vgpr5 def $vgpr5_vgpr6_vgpr7_vgpr8 killed $exec
; GFX9-O0-NEXT:    v_mov_b32_e32 v6, v11
; GFX9-O0-NEXT:    v_mov_b32_e32 v7, v10
; GFX9-O0-NEXT:    v_mov_b32_e32 v8, v9
; GFX9-O0-NEXT:    buffer_store_dwordx4 v[5:8], v0, s[0:3], s4 offen
; GFX9-O0-NEXT:    buffer_store_dwordx2 v[3:4], v0, s[0:3], s4 offen offset:16
; GFX9-O0-NEXT:    s_endpgm
;
; GFX9-O3-LABEL: strict_wwm_amdgpu_cs_main:
; GFX9-O3:       ; %bb.0:
; GFX9-O3-NEXT:    v_lshlrev_b32_e32 v0, 5, v0
; GFX9-O3-NEXT:    buffer_load_dwordx4 v[7:10], v0, s[0:3], 0 offen
; GFX9-O3-NEXT:    buffer_load_dwordx2 v[11:12], v0, s[0:3], 0 offen offset:16
; GFX9-O3-NEXT:    s_or_saveexec_b64 s[4:5], -1
; GFX9-O3-NEXT:    s_mov_b32 s6, -1
; GFX9-O3-NEXT:    s_brev_b32 s7, -2
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v2, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    s_waitcnt vmcnt(1)
; GFX9-O3-NEXT:    v_mov_b32_e32 v1, v7
; GFX9-O3-NEXT:    v_mov_b32_e32 v2, v8
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v3, v9
; GFX9-O3-NEXT:    v_mov_b32_e32 v4, v10
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, s6
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, s7
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    s_waitcnt vmcnt(0)
; GFX9-O3-NEXT:    v_mov_b32_e32 v5, v11
; GFX9-O3-NEXT:    v_mov_b32_e32 v6, v12
; GFX9-O3-NEXT:    s_mov_b64 exec, -1
; GFX9-O3-NEXT:    s_mov_b64 exec, s[4:5]
; GFX9-O3-NEXT:    v_mov_b32_e32 v7, v1
; GFX9-O3-NEXT:    v_mov_b32_e32 v9, v3
; GFX9-O3-NEXT:    v_mov_b32_e32 v8, v2
; GFX9-O3-NEXT:    v_mov_b32_e32 v10, v4
; GFX9-O3-NEXT:    v_mov_b32_e32 v11, v5
; GFX9-O3-NEXT:    v_mov_b32_e32 v12, v6
; GFX9-O3-NEXT:    buffer_store_dwordx4 v[7:10], v0, s[0:3], 0 offen
; GFX9-O3-NEXT:    buffer_store_dwordx2 v[11:12], v0, s[0:3], 0 offen offset:16
; GFX9-O3-NEXT:    s_endpgm
  %tmp17 = shl i32 %index, 5
  %tmp18 = tail call <4 x i32> @llvm.amdgcn.s.buffer.load.v4i32(<4 x i32> %desc, i32 %tmp17, i32 0)
  %.i0.upto1.bc = bitcast <4 x i32> %tmp18 to <2 x i64>
  %tmp19 = or i32 %tmp17, 16
  %tmp20 = tail call <2 x i32> @llvm.amdgcn.s.buffer.load.v2i32(<4 x i32> %desc, i32 %tmp19, i32 0)
  %.i0.upto1.extract = extractelement <2 x i64> %.i0.upto1.bc, i32 0
  %tmp22 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i0.upto1.extract, i64 9223372036854775807)
  %tmp97 = tail call i64 @llvm.amdgcn.strict.wwm.i64(i64 %tmp22)
  %.i1.upto1.extract = extractelement <2 x i64> %.i0.upto1.bc, i32 1
  %tmp99 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i1.upto1.extract, i64 9223372036854775807)
  %tmp174 = tail call i64 @llvm.amdgcn.strict.wwm.i64(i64 %tmp99)
  %.i25 = bitcast <2 x i32> %tmp20 to i64
  %tmp176 = tail call i64 @llvm.amdgcn.set.inactive.i64(i64 %.i25, i64 9223372036854775807)
  %tmp251 = tail call i64 @llvm.amdgcn.strict.wwm.i64(i64 %tmp176)
  %.cast = bitcast i64 %tmp97 to <2 x float>
  %.cast6 = bitcast i64 %tmp174 to <2 x float>
  %.cast7 = bitcast i64 %tmp251 to <2 x float>
  %tmp254 = shufflevector <2 x float> %.cast, <2 x float> %.cast6, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %desc.int = bitcast <4 x i32> %desc to i128
  %desc.ptr = inttoptr i128 %desc.int to ptr addrspace(8)
  tail call void @llvm.amdgcn.raw.ptr.buffer.store.v4f32(<4 x float> %tmp254, ptr addrspace(8) %desc.ptr, i32 %tmp17, i32 0, i32 0)
  tail call void @llvm.amdgcn.raw.ptr.buffer.store.v2f32(<2 x float> %.cast7, ptr addrspace(8)%desc.ptr, i32 %tmp19, i32 0, i32 0)
  ret void
}

declare i32 @llvm.amdgcn.strict.wwm.i32(i32)
declare i64 @llvm.amdgcn.strict.wwm.i64(i64)
declare i32 @llvm.amdgcn.wwm.i32(i32)
declare i64 @llvm.amdgcn.wwm.i64(i64)
declare i32 @llvm.amdgcn.set.inactive.i32(i32, i32)
declare i64 @llvm.amdgcn.set.inactive.i64(i64, i64)
declare i32 @llvm.amdgcn.update.dpp.i32(i32, i32, i32, i32, i32, i1)
declare <2 x float> @llvm.amdgcn.raw.ptr.buffer.load.v2f32(ptr addrspace(8), i32, i32, i32)
declare void @llvm.amdgcn.raw.ptr.buffer.store.f32(float, ptr addrspace(8), i32, i32, i32)
declare void @llvm.amdgcn.raw.ptr.buffer.store.i32(i32, ptr addrspace(8), i32, i32, i32)
declare void @llvm.amdgcn.raw.ptr.buffer.store.v2i32(<2 x i32>, ptr addrspace(8), i32, i32, i32)
declare void @llvm.amdgcn.raw.ptr.buffer.store.v2f32(<2 x float>, ptr addrspace(8), i32, i32, i32)
declare void @llvm.amdgcn.raw.ptr.buffer.store.v4f32(<4 x float>, ptr addrspace(8), i32, i32, i32)
declare <2 x i32> @llvm.amdgcn.s.buffer.load.v2i32(<4 x i32>, i32, i32)
declare <4 x i32> @llvm.amdgcn.s.buffer.load.v4i32(<4 x i32>, i32, i32)

!llvm.module.flags = !{!0}
!0 = !{i32 1, !"amdhsa_code_object_version", i32 500}