; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=amdgpu12.00 -global-isel=0 | FileCheck %s -check-prefix=GFX12-SDAG
; RUN: llc < %s -mtriple=amdgpu12.00 -global-isel=1 -global-isel-abort=2 | FileCheck %s -check-prefix=GFX12-GISEL

declare <2 x half> @llvm.amdgcn.struct.buffer.atomic.fadd.v2f16(<2 x half>, <4 x i32>, i32, i32, i32, i32 immarg)
declare <2 x bfloat> @llvm.amdgcn.struct.buffer.atomic.fadd.v2bf16(<2 x bfloat>, <4 x i32>, i32, i32, i32, i32 immarg)
declare <2 x half> @llvm.amdgcn.raw.buffer.atomic.fadd.v2f16(<2 x half>, <4 x i32>, i32, i32, i32)
declare <2 x bfloat> @llvm.amdgcn.raw.buffer.atomic.fadd.v2bf16(<2 x bfloat> %val, <4 x i32> %rsrc, i32, i32, i32)

define amdgpu_ps void @raw_buffer_atomic_add_v2f16_noret_offset(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2f16_noret_offset:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, off, s[0:3], s4 offset:92
; GFX12-SDAG-NEXT:    s_endpgm
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2f16_noret_offset:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, off, s[0:3], s4 offset:92
; GFX12-GISEL-NEXT:    s_endpgm
  %ret = call <2 x half> @llvm.amdgcn.raw.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 92, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @raw_buffer_atomic_add_v2f16_noret(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2f16_noret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, v1, s[0:3], s4 offen
; GFX12-SDAG-NEXT:    s_endpgm
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2f16_noret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, v1, s[0:3], s4 offen
; GFX12-GISEL-NEXT:    s_endpgm
  %ret = call <2 x half> @llvm.amdgcn.raw.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps <2 x half> @raw_buffer_atomic_add_v2f16_ret_offset(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2f16_ret_offset:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, off, s[0:3], s4 offset:92 th:TH_ATOMIC_RETURN
; GFX12-SDAG-NEXT:    s_wait_loadcnt 0x0
; GFX12-SDAG-NEXT:    ; return to shader part epilog
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2f16_ret_offset:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, off, s[0:3], s4 offset:92 th:TH_ATOMIC_RETURN
; GFX12-GISEL-NEXT:    s_wait_loadcnt 0x0
; GFX12-GISEL-NEXT:    ; return to shader part epilog
  %ret = call <2 x half> @llvm.amdgcn.raw.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 92, i32 %soffset, i32 0)
  ret <2 x half> %ret
}

define amdgpu_ps <2 x half> @raw_buffer_atomic_add_v2f16_ret(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2f16_ret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, v1, s[0:3], s4 offen th:TH_ATOMIC_RETURN
; GFX12-SDAG-NEXT:    s_wait_loadcnt 0x0
; GFX12-SDAG-NEXT:    ; return to shader part epilog
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2f16_ret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, v1, s[0:3], s4 offen th:TH_ATOMIC_RETURN
; GFX12-GISEL-NEXT:    s_wait_loadcnt 0x0
; GFX12-GISEL-NEXT:    ; return to shader part epilog
  %ret = call <2 x half> @llvm.amdgcn.raw.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret <2 x half> %ret
}

define amdgpu_ps float @struct_buffer_atomic_add_v2f16_ret(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: struct_buffer_atomic_add_v2f16_ret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, v[1:2], s[0:3], s4 idxen offen th:TH_ATOMIC_RETURN
; GFX12-SDAG-NEXT:    s_wait_loadcnt 0x0
; GFX12-SDAG-NEXT:    ; return to shader part epilog
;
; GFX12-GISEL-LABEL: struct_buffer_atomic_add_v2f16_ret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, v[1:2], s[0:3], s4 idxen offen th:TH_ATOMIC_RETURN
; GFX12-GISEL-NEXT:    s_wait_loadcnt 0x0
; GFX12-GISEL-NEXT:    ; return to shader part epilog
  %orig = call <2 x half> @llvm.amdgcn.struct.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  %r = bitcast <2 x half> %orig to float
  ret float %r
}

define amdgpu_ps void @struct_buffer_atomic_add_v2f16_noret(<2 x half> %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: struct_buffer_atomic_add_v2f16_noret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_f16 v0, v[1:2], s[0:3], s4 idxen offen
; GFX12-SDAG-NEXT:    s_endpgm
;
; GFX12-GISEL-LABEL: struct_buffer_atomic_add_v2f16_noret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_f16 v0, v[1:2], s[0:3], s4 idxen offen
; GFX12-GISEL-NEXT:    s_endpgm
  %orig = call <2 x half> @llvm.amdgcn.struct.buffer.atomic.fadd.v2f16(<2 x half> %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps float @struct_buffer_atomic_add_v2bf16_ret(<2 x bfloat> %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: struct_buffer_atomic_add_v2bf16_ret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_bf16 v0, v[1:2], s[0:3], s4 idxen offen th:TH_ATOMIC_RETURN
; GFX12-SDAG-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX12-SDAG-NEXT:    s_wait_loadcnt 0x0
; GFX12-SDAG-NEXT:    flat_store_b32 v[1:2], v0
; GFX12-SDAG-NEXT:    v_mov_b32_e32 v0, 1.0
; GFX12-SDAG-NEXT:    s_wait_dscnt 0x0
; GFX12-SDAG-NEXT:    ; return to shader part epilog
;
; GFX12-GISEL-LABEL: struct_buffer_atomic_add_v2bf16_ret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_bf16 v0, v[1:2], s[0:3], s4 idxen offen th:TH_ATOMIC_RETURN
; GFX12-GISEL-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX12-GISEL-NEXT:    s_wait_loadcnt 0x0
; GFX12-GISEL-NEXT:    flat_store_b32 v[1:2], v0
; GFX12-GISEL-NEXT:    v_mov_b32_e32 v0, 1.0
; GFX12-GISEL-NEXT:    s_wait_dscnt 0x0
; GFX12-GISEL-NEXT:    ; return to shader part epilog
  %orig = call <2 x bfloat> @llvm.amdgcn.struct.buffer.atomic.fadd.v2bf16(<2 x bfloat> %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  store <2 x bfloat> %orig, ptr null
  ret float 1.0
}

define amdgpu_ps void @struct_buffer_atomic_add_v2bf16_noret(<2 x bfloat> %val, <4 x i32> inreg %rsrc, i32 %vindex, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: struct_buffer_atomic_add_v2bf16_noret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_bf16 v0, v[1:2], s[0:3], s4 idxen offen
; GFX12-SDAG-NEXT:    s_endpgm
;
; GFX12-GISEL-LABEL: struct_buffer_atomic_add_v2bf16_noret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_bf16 v0, v[1:2], s[0:3], s4 idxen offen
; GFX12-GISEL-NEXT:    s_endpgm
  %orig = call <2 x bfloat> @llvm.amdgcn.struct.buffer.atomic.fadd.v2bf16(<2 x bfloat> %val, <4 x i32> %rsrc, i32 %vindex, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps void @raw_buffer_atomic_add_v2bf16(<2 x bfloat> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2bf16:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_bf16 v0, v1, s[0:3], s4 offen
; GFX12-SDAG-NEXT:    s_endpgm
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2bf16:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_bf16 v0, v1, s[0:3], s4 offen
; GFX12-GISEL-NEXT:    s_endpgm
  %ret = call <2 x bfloat> @llvm.amdgcn.raw.buffer.atomic.fadd.v2bf16(<2 x bfloat> %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  ret void
}

define amdgpu_ps float @raw_buffer_atomic_add_v2bf16_ret(<2 x bfloat> %val, <4 x i32> inreg %rsrc, i32 %voffset, i32 inreg %soffset) {
; GFX12-SDAG-LABEL: raw_buffer_atomic_add_v2bf16_ret:
; GFX12-SDAG:       ; %bb.0:
; GFX12-SDAG-NEXT:    buffer_atomic_pk_add_bf16 v0, v1, s[0:3], s4 offen th:TH_ATOMIC_RETURN
; GFX12-SDAG-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX12-SDAG-NEXT:    s_wait_loadcnt 0x0
; GFX12-SDAG-NEXT:    flat_store_b32 v[1:2], v0
; GFX12-SDAG-NEXT:    v_mov_b32_e32 v0, 1.0
; GFX12-SDAG-NEXT:    s_wait_dscnt 0x0
; GFX12-SDAG-NEXT:    ; return to shader part epilog
;
; GFX12-GISEL-LABEL: raw_buffer_atomic_add_v2bf16_ret:
; GFX12-GISEL:       ; %bb.0:
; GFX12-GISEL-NEXT:    buffer_atomic_pk_add_bf16 v0, v1, s[0:3], s4 offen th:TH_ATOMIC_RETURN
; GFX12-GISEL-NEXT:    v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, 0
; GFX12-GISEL-NEXT:    s_wait_loadcnt 0x0
; GFX12-GISEL-NEXT:    flat_store_b32 v[1:2], v0
; GFX12-GISEL-NEXT:    v_mov_b32_e32 v0, 1.0
; GFX12-GISEL-NEXT:    s_wait_dscnt 0x0
; GFX12-GISEL-NEXT:    ; return to shader part epilog
  %orig = call <2 x bfloat> @llvm.amdgcn.raw.buffer.atomic.fadd.v2bf16(<2 x bfloat> %val, <4 x i32> %rsrc, i32 %voffset, i32 %soffset, i32 0)
  store <2 x bfloat> %orig, ptr null
  ret float 1.0
}