# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx908 -verify-machineinstrs -run-pass=regbankselect -o - %s | FileCheck %s
# Make sure that an arbitrary AGPR is treated as a divergent value
# that needs to be copied to VGPR, and then waterfalled
# 32-bit case
---
name: raw_buffer_store__sgpr_rsrc__vgpr_val__vgpr_voffset__agpr_soffset
legalized: true
tracksRegLiveness: true
body: |
bb.0:
liveins: $sgpr0_sgpr1_sgpr2_sgpr3, $vgpr0, $vgpr1, $agpr0
; CHECK-LABEL: name: raw_buffer_store__sgpr_rsrc__vgpr_val__vgpr_voffset__agpr_soffset
; CHECK: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $sgpr0_sgpr1_sgpr2_sgpr3, $vgpr0, $vgpr1, $agpr0
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: %val:vgpr(s32) = COPY $vgpr0
; CHECK-NEXT: %rsrc:sgpr(<4 x s32>) = COPY $sgpr0_sgpr1_sgpr2_sgpr3
; CHECK-NEXT: %agpr:agpr(s32) = COPY $agpr0
; CHECK-NEXT: %voffset:vgpr(s32) = COPY $vgpr1
; CHECK-NEXT: %zero:sgpr(s32) = G_CONSTANT i32 0
; CHECK-NEXT: [[COPY:%[0-9]+]]:vgpr(s32) = COPY %zero(s32)
; CHECK-NEXT: [[DEF:%[0-9]+]]:sreg_64_xexec = IMPLICIT_DEF
; CHECK-NEXT: [[COPY1:%[0-9]+]]:vgpr_32(s32) = COPY %agpr(s32)
; CHECK-NEXT: [[S_MOV_B64_:%[0-9]+]]:sreg_64_xexec = S_MOV_B64 $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .1:
; CHECK-NEXT: successors: %bb.2(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[PHI:%[0-9]+]]:sreg_64_xexec = PHI [[DEF]], %bb.0, %9, %bb.2
; CHECK-NEXT: [[V_READFIRSTLANE_B32_:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[COPY1]](s32), implicit $exec
; CHECK-NEXT: [[ICMP:%[0-9]+]]:vcc(s1) = G_ICMP intpred(eq), [[V_READFIRSTLANE_B32_]](s32), [[COPY1]]
; CHECK-NEXT: [[INT:%[0-9]+]]:sreg_64_xexec(s64) = G_INTRINSIC intrinsic(@llvm.amdgcn.ballot), [[ICMP]](s1)
; CHECK-NEXT: [[S_AND_SAVEEXEC_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_SAVEEXEC_B64 killed [[INT]](s64), implicit-def $exec, implicit-def $scc, implicit $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .2:
; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.1(0x40000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: G_AMDGPU_BUFFER_STORE %val(s32), %rsrc(<4 x s32>), [[COPY]](s32), %voffset, [[V_READFIRSTLANE_B32_]], 0, 0, 0 :: (dereferenceable store (s32), addrspace 4)
; CHECK-NEXT: $exec = S_XOR_B64_term $exec, [[S_AND_SAVEEXEC_B64_]], implicit-def $scc
; CHECK-NEXT: SI_WATERFALL_LOOP %bb.1, implicit $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .3:
; CHECK-NEXT: successors: %bb.4(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: $exec = S_MOV_B64_term [[S_MOV_B64_]]
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .4:
; CHECK-NEXT: S_ENDPGM 0
%val:_(s32) = COPY $vgpr0
%rsrc:_(<4 x s32>) = COPY $sgpr0_sgpr1_sgpr2_sgpr3
%agpr:_(s32) = COPY $agpr0
%voffset:_(s32) = COPY $vgpr1
%zero:_(s32) = G_CONSTANT i32 0
G_AMDGPU_BUFFER_STORE %val, %rsrc, %zero, %voffset, %agpr, 0, 0, 0 :: (dereferenceable store (s32), addrspace 4)
S_ENDPGM 0
...
# Register tuple case
---
name: load_1d_vgpr_vaddr__agpr_srsrc
legalized: true
tracksRegLiveness: true
body: |
bb.0:
liveins: $agpr0_agpr1_agpr2_agpr3_agpr4_agpr5_agpr6_agpr7, $vgpr0
; CHECK-LABEL: name: load_1d_vgpr_vaddr__agpr_srsrc
; CHECK: successors: %bb.1(0x80000000)
; CHECK-NEXT: liveins: $agpr0_agpr1_agpr2_agpr3_agpr4_agpr5_agpr6_agpr7, $vgpr0
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[COPY:%[0-9]+]]:agpr(<8 x s32>) = COPY $agpr0_agpr1_agpr2_agpr3_agpr4_agpr5_agpr6_agpr7
; CHECK-NEXT: [[COPY1:%[0-9]+]]:vgpr(s32) = COPY $vgpr0
; CHECK-NEXT: [[DEF:%[0-9]+]]:sreg_64_xexec = IMPLICIT_DEF
; CHECK-NEXT: [[COPY2:%[0-9]+]]:vgpr(<8 x s32>) = COPY [[COPY]](<8 x s32>)
; CHECK-NEXT: [[S_MOV_B64_:%[0-9]+]]:sreg_64_xexec = S_MOV_B64 $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .1:
; CHECK-NEXT: successors: %bb.2(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[PHI:%[0-9]+]]:sreg_64_xexec = PHI [[DEF]], %bb.0, %6, %bb.2
; CHECK-NEXT: [[UV:%[0-9]+]]:vgpr_32(s32), [[UV1:%[0-9]+]]:vgpr_32(s32), [[UV2:%[0-9]+]]:vgpr_32(s32), [[UV3:%[0-9]+]]:vgpr_32(s32), [[UV4:%[0-9]+]]:vgpr_32(s32), [[UV5:%[0-9]+]]:vgpr_32(s32), [[UV6:%[0-9]+]]:vgpr_32(s32), [[UV7:%[0-9]+]]:vgpr_32(s32) = G_UNMERGE_VALUES [[COPY2]](<8 x s32>)
; CHECK-NEXT: [[V_READFIRSTLANE_B32_:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_1:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV1]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_2:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV2]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_3:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV3]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_4:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV4]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_5:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV5]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_6:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV6]](s32), implicit $exec
; CHECK-NEXT: [[V_READFIRSTLANE_B32_7:%[0-9]+]]:sreg_32(s32) = V_READFIRSTLANE_B32 [[UV7]](s32), implicit $exec
; CHECK-NEXT: [[BUILD_VECTOR:%[0-9]+]]:sgpr(<8 x s32>) = G_BUILD_VECTOR [[V_READFIRSTLANE_B32_]](s32), [[V_READFIRSTLANE_B32_1]](s32), [[V_READFIRSTLANE_B32_2]](s32), [[V_READFIRSTLANE_B32_3]](s32), [[V_READFIRSTLANE_B32_4]](s32), [[V_READFIRSTLANE_B32_5]](s32), [[V_READFIRSTLANE_B32_6]](s32), [[V_READFIRSTLANE_B32_7]](s32)
; CHECK-NEXT: [[UV8:%[0-9]+]]:vgpr(s64), [[UV9:%[0-9]+]]:vgpr(s64), [[UV10:%[0-9]+]]:vgpr(s64), [[UV11:%[0-9]+]]:vgpr(s64) = G_UNMERGE_VALUES [[COPY2]](<8 x s32>)
; CHECK-NEXT: [[UV12:%[0-9]+]]:sgpr(s64), [[UV13:%[0-9]+]]:sgpr(s64), [[UV14:%[0-9]+]]:sgpr(s64), [[UV15:%[0-9]+]]:sgpr(s64) = G_UNMERGE_VALUES [[BUILD_VECTOR]](<8 x s32>)
; CHECK-NEXT: [[ICMP:%[0-9]+]]:vcc(s1) = G_ICMP intpred(eq), [[UV12]](s64), [[UV8]]
; CHECK-NEXT: [[ICMP1:%[0-9]+]]:vcc(s1) = G_ICMP intpred(eq), [[UV13]](s64), [[UV9]]
; CHECK-NEXT: [[AND:%[0-9]+]]:vcc(s1) = G_AND [[ICMP]], [[ICMP1]]
; CHECK-NEXT: [[ICMP2:%[0-9]+]]:vcc(s1) = G_ICMP intpred(eq), [[UV14]](s64), [[UV10]]
; CHECK-NEXT: [[AND1:%[0-9]+]]:vcc(s1) = G_AND [[AND]], [[ICMP2]]
; CHECK-NEXT: [[ICMP3:%[0-9]+]]:vcc(s1) = G_ICMP intpred(eq), [[UV15]](s64), [[UV11]]
; CHECK-NEXT: [[AND2:%[0-9]+]]:vcc(s1) = G_AND [[AND1]], [[ICMP3]]
; CHECK-NEXT: [[INT:%[0-9]+]]:sreg_64_xexec(s64) = G_INTRINSIC intrinsic(@llvm.amdgcn.ballot), [[AND2]](s1)
; CHECK-NEXT: [[S_AND_SAVEEXEC_B64_:%[0-9]+]]:sreg_64_xexec = S_AND_SAVEEXEC_B64 killed [[INT]](s64), implicit-def $exec, implicit-def $scc, implicit $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .2:
; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.1(0x40000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: [[AMDGPU_INTRIN_IMAGE_LOAD:%[0-9]+]]:vgpr(<4 x s32>) = G_AMDGPU_INTRIN_IMAGE_LOAD intrinsic(@llvm.amdgcn.image.load.1d), 15, [[COPY1]](s32), [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable load (<4 x s32>))
; CHECK-NEXT: $exec = S_XOR_B64_term $exec, [[S_AND_SAVEEXEC_B64_]], implicit-def $scc
; CHECK-NEXT: SI_WATERFALL_LOOP %bb.1, implicit $exec
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .3:
; CHECK-NEXT: successors: %bb.4(0x80000000)
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: $exec = S_MOV_B64_term [[S_MOV_B64_]]
; CHECK-NEXT: {{ $}}
; CHECK-NEXT: .4:
; CHECK-NEXT: S_ENDPGM 0, implicit [[AMDGPU_INTRIN_IMAGE_LOAD]](<4 x s32>)
%0:_(<8 x s32>) = COPY $agpr0_agpr1_agpr2_agpr3_agpr4_agpr5_agpr6_agpr7
%1:_(s32) = COPY $vgpr0
%2:_(<4 x s32>) = G_AMDGPU_INTRIN_IMAGE_LOAD intrinsic(@llvm.amdgcn.image.load.1d), 15, %1(s32), %0(<8 x s32>), 0, 0, 0 :: (dereferenceable load (<4 x s32>))
S_ENDPGM 0, implicit %2
...