# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=fiji -O0 -run-pass=legalizer %s -o - | FileCheck %s --check-prefix=GFX8
# RUN: llc -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx906 -O0 -run-pass=legalizer %s -o - | FileCheck %s --check-prefix=GFX9

---
name: test_smulh_s32
body: |
  bb.0:
    liveins: $vgpr0, $vgpr1

    ; GFX8-LABEL: name: test_smulh_s32
    ; GFX8: liveins: $vgpr0, $vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX8-NEXT: [[SMULH:%[0-9]+]]:_(s32) = G_SMULH [[COPY]], [[COPY1]]
    ; GFX8-NEXT: $vgpr0 = COPY [[SMULH]](s32)
    ; GFX9-LABEL: name: test_smulh_s32
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX9-NEXT: [[SMULH:%[0-9]+]]:_(s32) = G_SMULH [[COPY]], [[COPY1]]
    ; GFX9-NEXT: $vgpr0 = COPY [[SMULH]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = COPY $vgpr1
    %2:_(s32) = G_SMULH %0, %1
    $vgpr0 = COPY %2
...

---
name: test_smulh_v2s32
body: |
  bb.0:
    liveins: $vgpr0_vgpr1, $vgpr2_vgpr3

    ; GFX8-LABEL: name: test_smulh_v2s32
    ; GFX8: liveins: $vgpr0_vgpr1, $vgpr2_vgpr3
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr0_vgpr1
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr2_vgpr3
    ; GFX8-NEXT: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY]](<2 x s32>)
    ; GFX8-NEXT: [[UV2:%[0-9]+]]:_(s32), [[UV3:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY1]](<2 x s32>)
    ; GFX8-NEXT: [[SMULH:%[0-9]+]]:_(s32) = G_SMULH [[UV]], [[UV2]]
    ; GFX8-NEXT: [[SMULH1:%[0-9]+]]:_(s32) = G_SMULH [[UV1]], [[UV3]]
    ; GFX8-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[SMULH]](s32), [[SMULH1]](s32)
    ; GFX8-NEXT: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    ; GFX9-LABEL: name: test_smulh_v2s32
    ; GFX9: liveins: $vgpr0_vgpr1, $vgpr2_vgpr3
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr0_vgpr1
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr2_vgpr3
    ; GFX9-NEXT: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY]](<2 x s32>)
    ; GFX9-NEXT: [[UV2:%[0-9]+]]:_(s32), [[UV3:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY1]](<2 x s32>)
    ; GFX9-NEXT: [[SMULH:%[0-9]+]]:_(s32) = G_SMULH [[UV]], [[UV2]]
    ; GFX9-NEXT: [[SMULH1:%[0-9]+]]:_(s32) = G_SMULH [[UV1]], [[UV3]]
    ; GFX9-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[SMULH]](s32), [[SMULH1]](s32)
    ; GFX9-NEXT: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    %0:_(<2 x s32>) = COPY $vgpr0_vgpr1
    %1:_(<2 x s32>) = COPY $vgpr2_vgpr3
    %2:_(<2 x s32>) = G_SMULH %0, %1
    $vgpr0_vgpr1 = COPY %2
...

---
name: test_smulh_s16
body: |
  bb.0:
    liveins: $vgpr0, $vgpr1

    ; GFX8-LABEL: name: test_smulh_s16
    ; GFX8: liveins: $vgpr0, $vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX8-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY]], 16
    ; GFX8-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY1]], 16
    ; GFX8-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG]], [[SEXT_INREG1]]
    ; GFX8-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX8-NEXT: [[ASHR:%[0-9]+]]:_(s32) = G_ASHR [[MUL]], [[C]](s32)
    ; GFX8-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[ASHR]], 16
    ; GFX8-NEXT: $vgpr0 = COPY [[SEXT_INREG2]](s32)
    ; GFX9-LABEL: name: test_smulh_s16
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX9-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY]], 16
    ; GFX9-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY1]], 16
    ; GFX9-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG]], [[SEXT_INREG1]]
    ; GFX9-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX9-NEXT: [[ASHR:%[0-9]+]]:_(s32) = G_ASHR [[MUL]], [[C]](s32)
    ; GFX9-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[ASHR]], 16
    ; GFX9-NEXT: $vgpr0 = COPY [[SEXT_INREG2]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = COPY $vgpr1
    %2:_(s16) = G_TRUNC %0
    %3:_(s16) = G_TRUNC %1
    %4:_(s16) = G_SMULH %2, %3
    %5:_(s32) = G_SEXT %4
    $vgpr0 = COPY %5
...

---
name: test_smulh_s8
body: |
  bb.0:
    liveins: $vgpr0, $vgpr1

    ; GFX8-LABEL: name: test_smulh_s8
    ; GFX8: liveins: $vgpr0, $vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX8-NEXT: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY]](s32)
    ; GFX8-NEXT: [[C:%[0-9]+]]:_(s16) = G_CONSTANT i16 8
    ; GFX8-NEXT: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[TRUNC]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[SHL]], [[C]](s16)
    ; GFX8-NEXT: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[COPY1]](s32)
    ; GFX8-NEXT: [[SHL1:%[0-9]+]]:_(s16) = G_SHL [[TRUNC1]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR1:%[0-9]+]]:_(s16) = G_ASHR [[SHL1]], [[C]](s16)
    ; GFX8-NEXT: [[MUL:%[0-9]+]]:_(s16) = G_MUL [[ASHR]], [[ASHR1]]
    ; GFX8-NEXT: [[ASHR2:%[0-9]+]]:_(s16) = G_ASHR [[MUL]], [[C]](s16)
    ; GFX8-NEXT: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR2]](s16)
    ; GFX8-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[ANYEXT]], 8
    ; GFX8-NEXT: $vgpr0 = COPY [[SEXT_INREG]](s32)
    ; GFX9-LABEL: name: test_smulh_s8
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX9-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY]], 8
    ; GFX9-NEXT: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[SEXT_INREG]](s32)
    ; GFX9-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY1]], 8
    ; GFX9-NEXT: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[SEXT_INREG1]](s32)
    ; GFX9-NEXT: [[MUL:%[0-9]+]]:_(s16) = G_MUL [[TRUNC]], [[TRUNC1]]
    ; GFX9-NEXT: [[C:%[0-9]+]]:_(s16) = G_CONSTANT i16 8
    ; GFX9-NEXT: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[MUL]], [[C]](s16)
    ; GFX9-NEXT: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR]](s16)
    ; GFX9-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[ANYEXT]], 8
    ; GFX9-NEXT: $vgpr0 = COPY [[SEXT_INREG2]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = COPY $vgpr1
    %2:_(s8) = G_TRUNC %0
    %3:_(s8) = G_TRUNC %1
    %4:_(s8) = G_SMULH %2, %3
    %5:_(s32) = G_SEXT %4
    $vgpr0 = COPY %5
...

---
name: test_smulh_v2s16
body: |
  bb.0:
    liveins: $vgpr0_vgpr1, $vgpr2_vgpr3
    ; GFX8-LABEL: name: test_smulh_v2s16
    ; GFX8: liveins: $vgpr0_vgpr1, $vgpr2_vgpr3
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr0_vgpr1
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr2_vgpr3
    ; GFX8-NEXT: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY]](<2 x s32>)
    ; GFX8-NEXT: [[UV2:%[0-9]+]]:_(s32), [[UV3:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY1]](<2 x s32>)
    ; GFX8-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV]], 16
    ; GFX8-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV2]], 16
    ; GFX8-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG]], [[SEXT_INREG1]]
    ; GFX8-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX8-NEXT: [[ASHR:%[0-9]+]]:_(s32) = G_ASHR [[MUL]], [[C]](s32)
    ; GFX8-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV1]], 16
    ; GFX8-NEXT: [[SEXT_INREG3:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV3]], 16
    ; GFX8-NEXT: [[MUL1:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG2]], [[SEXT_INREG3]]
    ; GFX8-NEXT: [[ASHR1:%[0-9]+]]:_(s32) = G_ASHR [[MUL1]], [[C]](s32)
    ; GFX8-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
    ; GFX8-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[ASHR]], [[C1]]
    ; GFX8-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[ASHR1]], [[C1]]
    ; GFX8-NEXT: [[SHL:%[0-9]+]]:_(s32) = G_SHL [[AND1]], [[C]](s32)
    ; GFX8-NEXT: [[OR:%[0-9]+]]:_(s32) = G_OR [[AND]], [[SHL]]
    ; GFX8-NEXT: [[BITCAST:%[0-9]+]]:_(<2 x s16>) = G_BITCAST [[OR]](s32)
    ; GFX8-NEXT: [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[BITCAST]](<2 x s16>)
    ; GFX8-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST1]], [[C]](s32)
    ; GFX8-NEXT: [[SEXT_INREG4:%[0-9]+]]:_(s32) = G_SEXT_INREG [[BITCAST1]], 16
    ; GFX8-NEXT: [[SEXT_INREG5:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR]], 16
    ; GFX8-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[SEXT_INREG4]](s32), [[SEXT_INREG5]](s32)
    ; GFX8-NEXT: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    ; GFX9-LABEL: name: test_smulh_v2s16
    ; GFX9: liveins: $vgpr0_vgpr1, $vgpr2_vgpr3
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr0_vgpr1
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr2_vgpr3
    ; GFX9-NEXT: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY]](<2 x s32>)
    ; GFX9-NEXT: [[UV2:%[0-9]+]]:_(s32), [[UV3:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY1]](<2 x s32>)
    ; GFX9-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV]], 16
    ; GFX9-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV2]], 16
    ; GFX9-NEXT: [[MUL:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG]], [[SEXT_INREG1]]
    ; GFX9-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX9-NEXT: [[ASHR:%[0-9]+]]:_(s32) = G_ASHR [[MUL]], [[C]](s32)
    ; GFX9-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV1]], 16
    ; GFX9-NEXT: [[SEXT_INREG3:%[0-9]+]]:_(s32) = G_SEXT_INREG [[UV3]], 16
    ; GFX9-NEXT: [[MUL1:%[0-9]+]]:_(s32) = G_MUL [[SEXT_INREG2]], [[SEXT_INREG3]]
    ; GFX9-NEXT: [[ASHR1:%[0-9]+]]:_(s32) = G_ASHR [[MUL1]], [[C]](s32)
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[ASHR]](s32), [[ASHR1]](s32)
    ; GFX9-NEXT: [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[BUILD_VECTOR_TRUNC]](<2 x s16>)
    ; GFX9-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
    ; GFX9-NEXT: [[SEXT_INREG4:%[0-9]+]]:_(s32) = G_SEXT_INREG [[BITCAST]], 16
    ; GFX9-NEXT: [[SEXT_INREG5:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR]], 16
    ; GFX9-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[SEXT_INREG4]](s32), [[SEXT_INREG5]](s32)
    ; GFX9-NEXT: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    %0:_(<2 x s32>) = COPY $vgpr0_vgpr1
    %1:_(<2 x s32>) = COPY $vgpr2_vgpr3
    %2:_(<2 x s16>) = G_TRUNC %0
    %3:_(<2 x s16>) = G_TRUNC %1
    %4:_(<2 x s16>) = G_SMULH %2, %3
    %5:_(<2 x s32>) = G_SEXT %4
    $vgpr0_vgpr1 = COPY %5
...

---
name: test_smulh_v2s8
body: |
  bb.0:
    liveins: $vgpr0, $vgpr1, $vgpr2, $vgpr3
    ; GFX8-LABEL: name: test_smulh_v2s8
    ; GFX8: liveins: $vgpr0, $vgpr1, $vgpr2, $vgpr3
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX8-NEXT: [[COPY2:%[0-9]+]]:_(s32) = COPY $vgpr2
    ; GFX8-NEXT: [[COPY3:%[0-9]+]]:_(s32) = COPY $vgpr3
    ; GFX8-NEXT: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY]](s32)
    ; GFX8-NEXT: [[C:%[0-9]+]]:_(s16) = G_CONSTANT i16 8
    ; GFX8-NEXT: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[TRUNC]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[SHL]], [[C]](s16)
    ; GFX8-NEXT: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[COPY2]](s32)
    ; GFX8-NEXT: [[SHL1:%[0-9]+]]:_(s16) = G_SHL [[TRUNC1]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR1:%[0-9]+]]:_(s16) = G_ASHR [[SHL1]], [[C]](s16)
    ; GFX8-NEXT: [[MUL:%[0-9]+]]:_(s16) = G_MUL [[ASHR]], [[ASHR1]]
    ; GFX8-NEXT: [[ASHR2:%[0-9]+]]:_(s16) = G_ASHR [[MUL]], [[C]](s16)
    ; GFX8-NEXT: [[TRUNC2:%[0-9]+]]:_(s16) = G_TRUNC [[COPY1]](s32)
    ; GFX8-NEXT: [[SHL2:%[0-9]+]]:_(s16) = G_SHL [[TRUNC2]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR3:%[0-9]+]]:_(s16) = G_ASHR [[SHL2]], [[C]](s16)
    ; GFX8-NEXT: [[TRUNC3:%[0-9]+]]:_(s16) = G_TRUNC [[COPY3]](s32)
    ; GFX8-NEXT: [[SHL3:%[0-9]+]]:_(s16) = G_SHL [[TRUNC3]], [[C]](s16)
    ; GFX8-NEXT: [[ASHR4:%[0-9]+]]:_(s16) = G_ASHR [[SHL3]], [[C]](s16)
    ; GFX8-NEXT: [[MUL1:%[0-9]+]]:_(s16) = G_MUL [[ASHR3]], [[ASHR4]]
    ; GFX8-NEXT: [[ASHR5:%[0-9]+]]:_(s16) = G_ASHR [[MUL1]], [[C]](s16)
    ; GFX8-NEXT: [[C1:%[0-9]+]]:_(s16) = G_CONSTANT i16 255
    ; GFX8-NEXT: [[AND:%[0-9]+]]:_(s16) = G_AND [[ASHR2]], [[C1]]
    ; GFX8-NEXT: [[AND1:%[0-9]+]]:_(s16) = G_AND [[ASHR5]], [[C1]]
    ; GFX8-NEXT: [[SHL4:%[0-9]+]]:_(s16) = G_SHL [[AND1]], [[C]](s16)
    ; GFX8-NEXT: [[OR:%[0-9]+]]:_(s16) = G_OR [[AND]], [[SHL4]]
    ; GFX8-NEXT: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[OR]](s16)
    ; GFX8-NEXT: $vgpr0 = COPY [[ANYEXT]](s32)
    ; GFX9-LABEL: name: test_smulh_v2s8
    ; GFX9: liveins: $vgpr0, $vgpr1, $vgpr2, $vgpr3
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX9-NEXT: [[COPY2:%[0-9]+]]:_(s32) = COPY $vgpr2
    ; GFX9-NEXT: [[COPY3:%[0-9]+]]:_(s32) = COPY $vgpr3
    ; GFX9-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY]], 8
    ; GFX9-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY1]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG]](s32), [[SEXT_INREG1]](s32)
    ; GFX9-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY2]], 8
    ; GFX9-NEXT: [[SEXT_INREG3:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY3]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC1:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG2]](s32), [[SEXT_INREG3]](s32)
    ; GFX9-NEXT: [[MUL:%[0-9]+]]:_(<2 x s16>) = G_MUL [[BUILD_VECTOR_TRUNC]], [[BUILD_VECTOR_TRUNC1]]
    ; GFX9-NEXT: [[C:%[0-9]+]]:_(s16) = G_CONSTANT i16 8
    ; GFX9-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
    ; GFX9-NEXT: [[COPY4:%[0-9]+]]:_(s32) = COPY [[C1]](s32)
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC2:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[COPY4]](s32), [[C1]](s32)
    ; GFX9-NEXT: [[ASHR:%[0-9]+]]:_(<2 x s16>) = G_ASHR [[MUL]], [[BUILD_VECTOR_TRUNC2]](<2 x s16>)
    ; GFX9-NEXT: [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[ASHR]](<2 x s16>)
    ; GFX9-NEXT: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[BITCAST]](s32)
    ; GFX9-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX9-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C2]](s32)
    ; GFX9-NEXT: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR]](s32)
    ; GFX9-NEXT: [[C3:%[0-9]+]]:_(s16) = G_CONSTANT i16 255
    ; GFX9-NEXT: [[AND:%[0-9]+]]:_(s16) = G_AND [[TRUNC]], [[C3]]
    ; GFX9-NEXT: [[AND1:%[0-9]+]]:_(s16) = G_AND [[TRUNC1]], [[C3]]
    ; GFX9-NEXT: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[AND1]], [[C]](s16)
    ; GFX9-NEXT: [[OR:%[0-9]+]]:_(s16) = G_OR [[AND]], [[SHL]]
    ; GFX9-NEXT: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[OR]](s16)
    ; GFX9-NEXT: $vgpr0 = COPY [[ANYEXT]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = COPY $vgpr1
    %2:_(s32) = COPY $vgpr2
    %3:_(s32) = COPY $vgpr3
    %5:_(s8) = G_TRUNC %0
    %6:_(s8) = G_TRUNC %1
    %7:_(s8) = G_TRUNC %2
    %8:_(s8) = G_TRUNC %3
    %11:_(<2 x s8>) = G_BUILD_VECTOR %5, %6
    %12:_(<2 x s8>) = G_BUILD_VECTOR %7, %8
    %13:_(<2 x s8>) = G_SMULH %11, %12
    %14:_(s8), %15:_(s8) = G_UNMERGE_VALUES %13
    %17:_(s16) = G_MERGE_VALUES %14, %15
    %18:_(s32) = G_ANYEXT %17
    $vgpr0 = COPY %18

...

---
name: test_smulh_v4s8
body: |
  bb.0:
    liveins: $vgpr0, $vgpr1
    ; GFX8-LABEL: name: test_smulh_v4s8
    ; GFX8: liveins: $vgpr0, $vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX8-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX8-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
    ; GFX8-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
    ; GFX8-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX8-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C1]](s32)
    ; GFX8-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
    ; GFX8-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C2]](s32)
    ; GFX8-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C]](s32)
    ; GFX8-NEXT: [[LSHR4:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C1]](s32)
    ; GFX8-NEXT: [[LSHR5:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C2]](s32)
    ; GFX8-NEXT: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY]](s32)
    ; GFX8-NEXT: [[C3:%[0-9]+]]:_(s16) = G_CONSTANT i16 8
    ; GFX8-NEXT: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[TRUNC]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[SHL]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[COPY1]](s32)
    ; GFX8-NEXT: [[SHL1:%[0-9]+]]:_(s16) = G_SHL [[TRUNC1]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR1:%[0-9]+]]:_(s16) = G_ASHR [[SHL1]], [[C3]](s16)
    ; GFX8-NEXT: [[MUL:%[0-9]+]]:_(s16) = G_MUL [[ASHR]], [[ASHR1]]
    ; GFX8-NEXT: [[ASHR2:%[0-9]+]]:_(s16) = G_ASHR [[MUL]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC2:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR]](s32)
    ; GFX8-NEXT: [[SHL2:%[0-9]+]]:_(s16) = G_SHL [[TRUNC2]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR3:%[0-9]+]]:_(s16) = G_ASHR [[SHL2]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC3:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR3]](s32)
    ; GFX8-NEXT: [[SHL3:%[0-9]+]]:_(s16) = G_SHL [[TRUNC3]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR4:%[0-9]+]]:_(s16) = G_ASHR [[SHL3]], [[C3]](s16)
    ; GFX8-NEXT: [[MUL1:%[0-9]+]]:_(s16) = G_MUL [[ASHR3]], [[ASHR4]]
    ; GFX8-NEXT: [[ASHR5:%[0-9]+]]:_(s16) = G_ASHR [[MUL1]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC4:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR1]](s32)
    ; GFX8-NEXT: [[SHL4:%[0-9]+]]:_(s16) = G_SHL [[TRUNC4]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR6:%[0-9]+]]:_(s16) = G_ASHR [[SHL4]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC5:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR4]](s32)
    ; GFX8-NEXT: [[SHL5:%[0-9]+]]:_(s16) = G_SHL [[TRUNC5]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR7:%[0-9]+]]:_(s16) = G_ASHR [[SHL5]], [[C3]](s16)
    ; GFX8-NEXT: [[MUL2:%[0-9]+]]:_(s16) = G_MUL [[ASHR6]], [[ASHR7]]
    ; GFX8-NEXT: [[ASHR8:%[0-9]+]]:_(s16) = G_ASHR [[MUL2]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC6:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR2]](s32)
    ; GFX8-NEXT: [[SHL6:%[0-9]+]]:_(s16) = G_SHL [[TRUNC6]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR9:%[0-9]+]]:_(s16) = G_ASHR [[SHL6]], [[C3]](s16)
    ; GFX8-NEXT: [[TRUNC7:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR5]](s32)
    ; GFX8-NEXT: [[SHL7:%[0-9]+]]:_(s16) = G_SHL [[TRUNC7]], [[C3]](s16)
    ; GFX8-NEXT: [[ASHR10:%[0-9]+]]:_(s16) = G_ASHR [[SHL7]], [[C3]](s16)
    ; GFX8-NEXT: [[MUL3:%[0-9]+]]:_(s16) = G_MUL [[ASHR9]], [[ASHR10]]
    ; GFX8-NEXT: [[ASHR11:%[0-9]+]]:_(s16) = G_ASHR [[MUL3]], [[C3]](s16)
    ; GFX8-NEXT: [[C4:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
    ; GFX8-NEXT: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR2]](s16)
    ; GFX8-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[ANYEXT]], [[C4]]
    ; GFX8-NEXT: [[ANYEXT1:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR5]](s16)
    ; GFX8-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[ANYEXT1]], [[C4]]
    ; GFX8-NEXT: [[SHL8:%[0-9]+]]:_(s32) = G_SHL [[AND1]], [[C]](s32)
    ; GFX8-NEXT: [[OR:%[0-9]+]]:_(s32) = G_OR [[AND]], [[SHL8]]
    ; GFX8-NEXT: [[ANYEXT2:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR8]](s16)
    ; GFX8-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[ANYEXT2]], [[C4]]
    ; GFX8-NEXT: [[SHL9:%[0-9]+]]:_(s32) = G_SHL [[AND2]], [[C1]](s32)
    ; GFX8-NEXT: [[OR1:%[0-9]+]]:_(s32) = G_OR [[OR]], [[SHL9]]
    ; GFX8-NEXT: [[ANYEXT3:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR11]](s16)
    ; GFX8-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[ANYEXT3]], [[C4]]
    ; GFX8-NEXT: [[SHL10:%[0-9]+]]:_(s32) = G_SHL [[AND3]], [[C2]](s32)
    ; GFX8-NEXT: [[OR2:%[0-9]+]]:_(s32) = G_OR [[OR1]], [[SHL10]]
    ; GFX8-NEXT: $vgpr0 = COPY [[OR2]](s32)
    ; GFX9-LABEL: name: test_smulh_v4s8
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
    ; GFX9-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
    ; GFX9-NEXT: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C]](s32)
    ; GFX9-NEXT: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; GFX9-NEXT: [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C1]](s32)
    ; GFX9-NEXT: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
    ; GFX9-NEXT: [[LSHR2:%[0-9]+]]:_(s32) = G_LSHR [[COPY]], [[C2]](s32)
    ; GFX9-NEXT: [[LSHR3:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C]](s32)
    ; GFX9-NEXT: [[LSHR4:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C1]](s32)
    ; GFX9-NEXT: [[LSHR5:%[0-9]+]]:_(s32) = G_LSHR [[COPY1]], [[C2]](s32)
    ; GFX9-NEXT: [[SEXT_INREG:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY]], 8
    ; GFX9-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG]](s32), [[SEXT_INREG1]](s32)
    ; GFX9-NEXT: [[SEXT_INREG2:%[0-9]+]]:_(s32) = G_SEXT_INREG [[COPY1]], 8
    ; GFX9-NEXT: [[SEXT_INREG3:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR3]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC1:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG2]](s32), [[SEXT_INREG3]](s32)
    ; GFX9-NEXT: [[MUL:%[0-9]+]]:_(<2 x s16>) = G_MUL [[BUILD_VECTOR_TRUNC]], [[BUILD_VECTOR_TRUNC1]]
    ; GFX9-NEXT: [[COPY2:%[0-9]+]]:_(s32) = COPY [[C]](s32)
    ; GFX9-NEXT: [[COPY3:%[0-9]+]]:_(s32) = COPY [[C]](s32)
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC2:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[COPY2]](s32), [[COPY3]](s32)
    ; GFX9-NEXT: [[ASHR:%[0-9]+]]:_(<2 x s16>) = G_ASHR [[MUL]], [[BUILD_VECTOR_TRUNC2]](<2 x s16>)
    ; GFX9-NEXT: [[SEXT_INREG4:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR1]], 8
    ; GFX9-NEXT: [[SEXT_INREG5:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR2]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC3:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG4]](s32), [[SEXT_INREG5]](s32)
    ; GFX9-NEXT: [[SEXT_INREG6:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR4]], 8
    ; GFX9-NEXT: [[SEXT_INREG7:%[0-9]+]]:_(s32) = G_SEXT_INREG [[LSHR5]], 8
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC4:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[SEXT_INREG6]](s32), [[SEXT_INREG7]](s32)
    ; GFX9-NEXT: [[MUL1:%[0-9]+]]:_(<2 x s16>) = G_MUL [[BUILD_VECTOR_TRUNC3]], [[BUILD_VECTOR_TRUNC4]]
    ; GFX9-NEXT: [[COPY4:%[0-9]+]]:_(s32) = COPY [[C]](s32)
    ; GFX9-NEXT: [[COPY5:%[0-9]+]]:_(s32) = COPY [[C]](s32)
    ; GFX9-NEXT: [[BUILD_VECTOR_TRUNC5:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR_TRUNC [[COPY4]](s32), [[COPY5]](s32)
    ; GFX9-NEXT: [[ASHR1:%[0-9]+]]:_(<2 x s16>) = G_ASHR [[MUL1]], [[BUILD_VECTOR_TRUNC5]](<2 x s16>)
    ; GFX9-NEXT: [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[ASHR]](<2 x s16>)
    ; GFX9-NEXT: [[LSHR6:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C1]](s32)
    ; GFX9-NEXT: [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[ASHR1]](<2 x s16>)
    ; GFX9-NEXT: [[LSHR7:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST1]], [[C1]](s32)
    ; GFX9-NEXT: [[C3:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
    ; GFX9-NEXT: [[AND:%[0-9]+]]:_(s32) = G_AND [[BITCAST]], [[C3]]
    ; GFX9-NEXT: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR6]], [[C3]]
    ; GFX9-NEXT: [[SHL:%[0-9]+]]:_(s32) = G_SHL [[AND1]], [[C]](s32)
    ; GFX9-NEXT: [[OR:%[0-9]+]]:_(s32) = G_OR [[AND]], [[SHL]]
    ; GFX9-NEXT: [[AND2:%[0-9]+]]:_(s32) = G_AND [[BITCAST1]], [[C3]]
    ; GFX9-NEXT: [[SHL1:%[0-9]+]]:_(s32) = G_SHL [[AND2]], [[C1]](s32)
    ; GFX9-NEXT: [[OR1:%[0-9]+]]:_(s32) = G_OR [[OR]], [[SHL1]]
    ; GFX9-NEXT: [[AND3:%[0-9]+]]:_(s32) = G_AND [[LSHR7]], [[C3]]
    ; GFX9-NEXT: [[SHL2:%[0-9]+]]:_(s32) = G_SHL [[AND3]], [[C2]](s32)
    ; GFX9-NEXT: [[OR2:%[0-9]+]]:_(s32) = G_OR [[OR1]], [[SHL2]]
    ; GFX9-NEXT: $vgpr0 = COPY [[OR2]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = COPY $vgpr1
    %2:_(s8), %3:_(s8), %4:_(s8), %5:_(s8) = G_UNMERGE_VALUES %0
    %6:_(s8), %7:_(s8), %8:_(s8), %9:_(s8) = G_UNMERGE_VALUES %1
    %10:_(<4 x s8>) = G_BUILD_VECTOR %2:_(s8), %3:_(s8), %4:_(s8), %5:_(s8)
    %11:_(<4 x s8>) = G_BUILD_VECTOR %6:_(s8), %7:_(s8), %8:_(s8), %9:_(s8)
    %12:_(<4 x s8>) = G_SMULH %10:_, %11:_
    %13:_(s8), %14:_(s8), %15:_(s8), %16:_(s8) = G_UNMERGE_VALUES %12:_(<4 x s8>)
    %17:_(s32) = G_MERGE_VALUES %13, %14, %15, %16
    $vgpr0 = COPY %17
...