Mercurial > hg > CbC > CbC_llvm

view llvm/test/CodeGen/AMDGPU/buffer-atomic-fadd.f64.ll @ 266:00f31e85ec16 default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Added tag current for changeset 31d058e83c98
author Shinji KONO <kono@ie.u-ryukyu.ac.jp>
date Sat, 14 Oct 2023 10:13:55 +0900
parents 1f2b6ac9f198
children
line wrap: on
line source

; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs -stop-after=amdgpu-isel < %s | FileCheck -check-prefix=GFX90A_GFX940 %s
; RUN: llc -march=amdgcn -mcpu=gfx940 -verify-machineinstrs -stop-after=amdgpu-isel < %s | FileCheck -check-prefix=GFX90A_GFX940 %s

define amdgpu_ps void @buffer_atomic_fadd_f64_offset_no_rtn(double %val, <4 x i32> inreg %rsrc, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_offset_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY1]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_OFFSET killed [[COPY7]], killed [[REG_SEQUENCE]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 0, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_atomic_fadd_f64_offen_no_rtn(double %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_offen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1, [[COPY3]], %subreg.sub2, [[COPY2]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_OFFEN killed [[COPY8]], [[COPY1]], killed [[REG_SEQUENCE]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_atomic_fadd_f64_idxen_no_rtn(double %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_idxen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1, [[COPY3]], %subreg.sub2, [[COPY2]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_IDXEN killed [[COPY8]], [[COPY1]], killed [[REG_SEQUENCE]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_atomic_fadd_f64_bothen_no_rtn(double %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_bothen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $vgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1, [[COPY4]], %subreg.sub2, [[COPY3]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY8]], %subreg.sub0, [[COPY7]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:vreg_64_align2 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_BOTHEN killed [[COPY9]], killed [[REG_SEQUENCE2]], killed [[REG_SEQUENCE]], [[COPY]], 0, 2, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 2)
  ret void
}

define amdgpu_ps double @buffer_atomic_fadd_f64_offset_rtn(double %val, <4 x i32> inreg %rsrc, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_offset_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1, [[COPY2]], %subreg.sub2, [[COPY1]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_OFFSET_RTN [[COPY7]], killed [[REG_SEQUENCE]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY8]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY9]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 0, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_atomic_fadd_f64_offen_rtn(double %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_offen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1, [[COPY3]], %subreg.sub2, [[COPY2]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_OFFEN_RTN [[COPY8]], [[COPY1]], killed [[REG_SEQUENCE]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY9]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY10]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_atomic_fadd_f64_idxen_rtn(double %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_idxen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1, [[COPY3]], %subreg.sub2, [[COPY2]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_IDXEN_RTN [[COPY8]], [[COPY1]], killed [[REG_SEQUENCE]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY9]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY10]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %vindex, i32 0, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_atomic_fadd_f64_bothen_rtn(double %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_atomic_fadd_f64_bothen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $vgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1, [[COPY4]], %subreg.sub2, [[COPY3]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY8]], %subreg.sub0, [[COPY7]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:vreg_64_align2 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE1]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN [[COPY9]], killed [[REG_SEQUENCE2]], killed [[REG_SEQUENCE]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64), align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY10]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY11]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps void @buffer_ptr_atomic_fadd_f64_offset_no_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_offset_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY10]], %subreg.sub0, killed [[COPY9]], %subreg.sub1, killed [[COPY8]], %subreg.sub2, killed [[COPY7]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_OFFSET killed [[COPY11]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 0, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_ptr_atomic_fadd_f64_offen_no_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_offen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY3]], %subreg.sub0, [[COPY2]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY11]], %subreg.sub0, killed [[COPY10]], %subreg.sub1, killed [[COPY9]], %subreg.sub2, killed [[COPY8]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_OFFEN killed [[COPY12]], [[COPY1]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_ptr_atomic_fadd_f64_idxen_no_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %vindex, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_idxen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY3]], %subreg.sub0, [[COPY2]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY11]], %subreg.sub0, killed [[COPY10]], %subreg.sub1, killed [[COPY9]], %subreg.sub2, killed [[COPY8]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_IDXEN killed [[COPY12]], [[COPY1]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 0, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @buffer_ptr_atomic_fadd_f64_bothen_no_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_bothen_no_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $vgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY12]], %subreg.sub0, killed [[COPY11]], %subreg.sub1, killed [[COPY10]], %subreg.sub2, killed [[COPY9]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY8]], %subreg.sub0, [[COPY7]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE4:%[0-9]+]]:vreg_64_align2 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY13:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   BUFFER_ATOMIC_ADD_F64_BOTHEN killed [[COPY13]], killed [[REG_SEQUENCE4]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 2, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   S_ENDPGM 0
  %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 2)
  ret void
}

define amdgpu_ps double @buffer_ptr_atomic_fadd_f64_offset_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_offset_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY10]], %subreg.sub0, killed [[COPY9]], %subreg.sub1, killed [[COPY8]], %subreg.sub2, killed [[COPY7]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_OFFSET_RTN [[COPY11]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY13:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFSET_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY12]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY13]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 0, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_ptr_atomic_fadd_f64_offen_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_offen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY3]], %subreg.sub0, [[COPY2]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY11]], %subreg.sub0, killed [[COPY10]], %subreg.sub1, killed [[COPY9]], %subreg.sub2, killed [[COPY8]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_OFFEN_RTN [[COPY12]], [[COPY1]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY13:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY14:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_OFFEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY13]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY14]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_ptr_atomic_fadd_f64_idxen_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %vindex, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_idxen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY3]], %subreg.sub0, [[COPY2]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY5]], %subreg.sub0, [[COPY4]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY11]], %subreg.sub0, killed [[COPY10]], %subreg.sub1, killed [[COPY9]], %subreg.sub2, killed [[COPY8]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY7]], %subreg.sub0, [[COPY6]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_IDXEN_RTN [[COPY12]], [[COPY1]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY13:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY14:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_IDXEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY13]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY14]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %vindex, i32 0, i32 %soffset, i32 0)
  ret double %ret
}

define amdgpu_ps double @buffer_ptr_atomic_fadd_f64_bothen_rtn(double %val, ptr addrspace(8) inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
  ; GFX90A_GFX940-LABEL: name: buffer_ptr_atomic_fadd_f64_bothen_rtn
  ; GFX90A_GFX940: bb.0 (%ir-block.0):
  ; GFX90A_GFX940-NEXT:   liveins: $vgpr0, $vgpr1, $sgpr0, $sgpr1, $sgpr2, $sgpr3, $vgpr2, $vgpr3, $sgpr4
  ; GFX90A_GFX940-NEXT: {{  $}}
  ; GFX90A_GFX940-NEXT:   [[COPY:%[0-9]+]]:sgpr_32 = COPY $sgpr4
  ; GFX90A_GFX940-NEXT:   [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY3:%[0-9]+]]:sgpr_32 = COPY $sgpr3
  ; GFX90A_GFX940-NEXT:   [[COPY4:%[0-9]+]]:sgpr_32 = COPY $sgpr2
  ; GFX90A_GFX940-NEXT:   [[COPY5:%[0-9]+]]:sgpr_32 = COPY $sgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY6:%[0-9]+]]:sgpr_32 = COPY $sgpr0
  ; GFX90A_GFX940-NEXT:   [[COPY7:%[0-9]+]]:vgpr_32 = COPY $vgpr1
  ; GFX90A_GFX940-NEXT:   [[COPY8:%[0-9]+]]:vgpr_32 = COPY $vgpr0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY4]], %subreg.sub0, [[COPY3]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY9:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY10:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE1:%[0-9]+]]:sgpr_64 = REG_SEQUENCE [[COPY6]], %subreg.sub0, [[COPY5]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY11:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub1
  ; GFX90A_GFX940-NEXT:   [[COPY12:%[0-9]+]]:sreg_32 = COPY [[REG_SEQUENCE1]].sub0
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE2:%[0-9]+]]:sgpr_128 = REG_SEQUENCE killed [[COPY12]], %subreg.sub0, killed [[COPY11]], %subreg.sub1, killed [[COPY10]], %subreg.sub2, killed [[COPY9]], %subreg.sub3
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE3:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[COPY8]], %subreg.sub0, [[COPY7]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[REG_SEQUENCE4:%[0-9]+]]:vreg_64_align2 = REG_SEQUENCE [[COPY2]], %subreg.sub0, [[COPY1]], %subreg.sub1
  ; GFX90A_GFX940-NEXT:   [[COPY13:%[0-9]+]]:vreg_64_align2 = COPY [[REG_SEQUENCE3]]
  ; GFX90A_GFX940-NEXT:   [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN:%[0-9]+]]:vreg_64_align2 = BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN [[COPY13]], killed [[REG_SEQUENCE4]], killed [[REG_SEQUENCE2]], [[COPY]], 0, 1, implicit $exec :: (volatile dereferenceable load store (s64) on %ir.rsrc, align 1, addrspace 8)
  ; GFX90A_GFX940-NEXT:   [[COPY14:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN]].sub0
  ; GFX90A_GFX940-NEXT:   [[COPY15:%[0-9]+]]:vgpr_32 = COPY [[BUFFER_ATOMIC_ADD_F64_BOTHEN_RTN]].sub1
  ; GFX90A_GFX940-NEXT:   $sgpr0 = COPY [[COPY14]]
  ; GFX90A_GFX940-NEXT:   $sgpr1 = COPY [[COPY15]]
  ; GFX90A_GFX940-NEXT:   SI_RETURN_TO_EPILOG $sgpr0, $sgpr1
  %ret = call double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double %val, ptr addrspace(8) %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  ret double %ret
}

declare double @llvm.amdgcn.raw.buffer.atomic.fadd.f64(double, <4 x i32>, i32, i32, i32 immarg)
declare double @llvm.amdgcn.struct.buffer.atomic.fadd.f64(double, <4 x i32>, i32, i32, i32, i32 immarg)

declare double @llvm.amdgcn.raw.ptr.buffer.atomic.fadd.f64(double, ptr addrspace(8), i32, i32, i32 immarg)
declare double @llvm.amdgcn.struct.ptr.buffer.atomic.fadd.f64(double, ptr addrspace(8), i32, i32, i32, i32 immarg)