llvm/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-fmul.v2s16.mir

# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -mtriple=amdgcn -mcpu=gfx900 -run-pass=instruction-select -global-isel-abort=0 -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX9 %s
# RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -run-pass=instruction-select -global-isel-abort=0 -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX9 %s
# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -run-pass=instruction-select -global-isel-abort=0 -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX9 %s

---
name: fmul_v2f16_vv
legalized: true
regBankSelected: true

body: |
  bb.0:
    liveins: $vgpr0, $vgpr1

    ; GFX9-LABEL: name: fmul_v2f16_vv
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
    ; GFX9-NEXT: %2:vgpr_32 = nofpexcept V_PK_MUL_F16 8, [[COPY]], 8, [[COPY1]], 0, 0, 0, 0, 0, implicit $mode, implicit $exec
    ; GFX9-NEXT: S_ENDPGM 0, implicit %2
    %0:vgpr(<2 x s16>) = COPY $vgpr0
    %1:vgpr(<2 x s16>) = COPY $vgpr1
    %2:vgpr(<2 x s16>) = G_FMUL %0, %1
    S_ENDPGM 0, implicit %2
...

---
name: fmul_v2f16_fneg_v_fneg_v
legalized: true
regBankSelected: true

body: |
  bb.0:
    liveins: $vgpr0, $vgpr1

    ; GFX9-LABEL: name: fmul_v2f16_fneg_v_fneg_v
    ; GFX9: liveins: $vgpr0, $vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
    ; GFX9-NEXT: %4:vgpr_32 = nofpexcept V_PK_MUL_F16 11, [[COPY]], 11, [[COPY1]], 0, 0, 0, 0, 0, implicit $mode, implicit $exec
    ; GFX9-NEXT: S_ENDPGM 0, implicit %4
    %0:vgpr(<2 x s16>) = COPY $vgpr0
    %1:vgpr(<2 x s16>) = COPY $vgpr1
    %2:vgpr(<2 x s16>) = G_FNEG %0
    %3:vgpr(<2 x s16>) = G_FNEG %1
    %4:vgpr(<2 x s16>) = G_FMUL %2, %3
    S_ENDPGM 0, implicit %4
...

---
name: fmul_v2f16_fneg_lo_v_v
legalized: true
regBankSelected: true

body: |
  bb.0:
    liveins: $vgpr0, $vgpr1, $vgpr2

    ; GFX9-LABEL: name: fmul_v2f16_fneg_lo_v_v
    ; GFX9: liveins: $vgpr0, $vgpr1, $vgpr2
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
    ; GFX9-NEXT: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
    ; GFX9-NEXT: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
    ; GFX9-NEXT: [[S_MOV_B32_:%[0-9]+]]:sreg_32 = S_MOV_B32 32768
    ; GFX9-NEXT: [[V_XOR_B32_e64_:%[0-9]+]]:vgpr_32 = V_XOR_B32_e64 [[S_MOV_B32_]], [[COPY1]], implicit $exec
    ; GFX9-NEXT: [[V_AND_B32_e32_:%[0-9]+]]:vgpr_32 = V_AND_B32_e32 65535, [[V_XOR_B32_e64_]], implicit $exec
    ; GFX9-NEXT: [[V_LSHL_OR_B32_e64_:%[0-9]+]]:vgpr_32 = V_LSHL_OR_B32_e64 [[COPY2]], 16, [[V_AND_B32_e32_]], implicit $exec
    ; GFX9-NEXT: %7:vgpr_32 = nofpexcept V_PK_MUL_F16 8, [[V_LSHL_OR_B32_e64_]], 8, [[COPY]], 0, 0, 0, 0, 0, implicit $mode, implicit $exec
    ; GFX9-NEXT: S_ENDPGM 0, implicit %7
    %0:vgpr(<2 x s16>) = COPY $vgpr0
    %1:vgpr(s32) = COPY $vgpr1
    %2:vgpr(s32) = COPY $vgpr2
    %3:vgpr(s16) = G_TRUNC %1
    %4:vgpr(s16) = G_FNEG %3
    %5:vgpr(s32) = G_ANYEXT %4
    %6:vgpr(<2 x s16>) = G_BUILD_VECTOR_TRUNC %5, %2
    %7:vgpr(<2 x s16>) = G_FMUL %6, %0
    S_ENDPGM 0, implicit %7
...