llvm/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-load-global.s96.mir

# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=hawaii -run-pass=instruction-select -verify-machineinstrs  -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX7 %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=hawaii -mattr=+flat-for-global -run-pass=instruction-select -verify-machineinstrs  -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX7-FLAT %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=fiji -run-pass=instruction-select -verify-machineinstrs -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX8 %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=gfx900 -run-pass=instruction-select -verify-machineinstrs -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX9 %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=gfx1010 -run-pass=instruction-select -verify-machineinstrs -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX10 %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=gfx1100 -run-pass=instruction-select -verify-machineinstrs -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX10 %s
# RUN: llc -amdgpu-global-isel-new-legality -mtriple=amdgcn -mcpu=gfx1200 -run-pass=instruction-select -verify-machineinstrs -global-isel-abort=0 -o - %s | FileCheck -check-prefix=GFX10 %s

---

name: load_global_v3s32
legalized:       true
regBankSelected: true
tracksRegLiveness: true

body: |
  bb.0:
    liveins:  $vgpr0_vgpr1

    ; GFX7-LABEL: name: load_global_v3s32
    ; GFX7: liveins: $vgpr0_vgpr1
    ; GFX7-NEXT: {{  $}}
    ; GFX7-NEXT: [[COPY:%[0-9]+]]:vreg_64 = COPY $vgpr0_vgpr1
    ; GFX7-NEXT: [[S_MOV_B32_:%[0-9]+]]:sreg_32 = S_MOV_B32 0
    ; GFX7-NEXT: [[S_MOV_B32_1:%[0-9]+]]:sreg_32 = S_MOV_B32 61440
    ; GFX7-NEXT: [[REG_SEQUENCE:%[0-9]+]]:sreg_64 = REG_SEQUENCE [[S_MOV_B32_]], %subreg.sub0, [[S_MOV_B32_1]], %subreg.sub1
    ; GFX7-NEXT: [[S_MOV_B64_:%[0-9]+]]:sreg_64 = S_MOV_B64 0
    ; GFX7-NEXT: [[REG_SEQUENCE1:%[0-9]+]]:sgpr_128 = REG_SEQUENCE [[S_MOV_B64_]], %subreg.sub0_sub1, [[REG_SEQUENCE]], %subreg.sub2_sub3
    ; GFX7-NEXT: [[BUFFER_LOAD_DWORDX3_ADDR64_:%[0-9]+]]:vreg_96 = BUFFER_LOAD_DWORDX3_ADDR64 [[COPY]], [[REG_SEQUENCE1]], 0, 0, 0, 0, implicit $exec :: (load (<3 x s32>), align 4, addrspace 1)
    ; GFX7-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[BUFFER_LOAD_DWORDX3_ADDR64_]]
    ;
    ; GFX7-FLAT-LABEL: name: load_global_v3s32
    ; GFX7-FLAT: liveins: $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: {{  $}}
    ; GFX7-FLAT-NEXT: [[COPY:%[0-9]+]]:vreg_64 = COPY $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: [[FLAT_LOAD_DWORDX3_:%[0-9]+]]:vreg_96 = FLAT_LOAD_DWORDX3 [[COPY]], 0, 0, implicit $exec, implicit $flat_scr :: (load (<3 x s32>), align 4, addrspace 1)
    ; GFX7-FLAT-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[FLAT_LOAD_DWORDX3_]]
    ;
    ; GFX8-LABEL: name: load_global_v3s32
    ; GFX8: liveins: $vgpr0_vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:vreg_64 = COPY $vgpr0_vgpr1
    ; GFX8-NEXT: [[FLAT_LOAD_DWORDX3_:%[0-9]+]]:vreg_96 = FLAT_LOAD_DWORDX3 [[COPY]], 0, 0, implicit $exec, implicit $flat_scr :: (load (<3 x s32>), align 4, addrspace 1)
    ; GFX8-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[FLAT_LOAD_DWORDX3_]]
    ;
    ; GFX9-LABEL: name: load_global_v3s32
    ; GFX9: liveins: $vgpr0_vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vreg_64 = COPY $vgpr0_vgpr1
    ; GFX9-NEXT: [[GLOBAL_LOAD_DWORDX3_:%[0-9]+]]:vreg_96 = GLOBAL_LOAD_DWORDX3 [[COPY]], 0, 0, implicit $exec :: (load (<3 x s32>), align 4, addrspace 1)
    ; GFX9-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[GLOBAL_LOAD_DWORDX3_]]
    ;
    ; GFX10-LABEL: name: load_global_v3s32
    ; GFX10: liveins: $vgpr0_vgpr1
    ; GFX10-NEXT: {{  $}}
    ; GFX10-NEXT: [[COPY:%[0-9]+]]:vreg_64 = COPY $vgpr0_vgpr1
    ; GFX10-NEXT: [[GLOBAL_LOAD_DWORDX3_:%[0-9]+]]:vreg_96 = GLOBAL_LOAD_DWORDX3 [[COPY]], 0, 0, implicit $exec :: (load (<3 x s32>), align 4, addrspace 1)
    ; GFX10-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[GLOBAL_LOAD_DWORDX3_]]
    %0:vgpr(p1) = COPY $vgpr0_vgpr1
    %1:vgpr(<3 x s32>) = G_LOAD %0 :: (load (<3 x s32>), align 4, addrspace 1)
    $vgpr0_vgpr1_vgpr2 = COPY %1

...

---

name: load_global_s96
legalized:       true
regBankSelected: true
tracksRegLiveness: true

body: |
  bb.0:
    liveins:  $vgpr0_vgpr1

    ; GFX7-LABEL: name: load_global_s96
    ; GFX7: liveins: $vgpr0_vgpr1
    ; GFX7-NEXT: {{  $}}
    ; GFX7-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX7-NEXT: [[LOAD:%[0-9]+]]:vreg_96(s96) = G_LOAD [[COPY]](p1) :: (load (s96), align 4, addrspace 1)
    ; GFX7-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](s96)
    ;
    ; GFX7-FLAT-LABEL: name: load_global_s96
    ; GFX7-FLAT: liveins: $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: {{  $}}
    ; GFX7-FLAT-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: [[LOAD:%[0-9]+]]:vreg_96(s96) = G_LOAD [[COPY]](p1) :: (load (s96), align 4, addrspace 1)
    ; GFX7-FLAT-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](s96)
    ;
    ; GFX8-LABEL: name: load_global_s96
    ; GFX8: liveins: $vgpr0_vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX8-NEXT: [[LOAD:%[0-9]+]]:vreg_96(s96) = G_LOAD [[COPY]](p1) :: (load (s96), align 4, addrspace 1)
    ; GFX8-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](s96)
    ;
    ; GFX9-LABEL: name: load_global_s96
    ; GFX9: liveins: $vgpr0_vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX9-NEXT: [[LOAD:%[0-9]+]]:vreg_96(s96) = G_LOAD [[COPY]](p1) :: (load (s96), align 4, addrspace 1)
    ; GFX9-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](s96)
    ;
    ; GFX10-LABEL: name: load_global_s96
    ; GFX10: liveins: $vgpr0_vgpr1
    ; GFX10-NEXT: {{  $}}
    ; GFX10-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX10-NEXT: [[LOAD:%[0-9]+]]:vreg_96(s96) = G_LOAD [[COPY]](p1) :: (load (s96), align 4, addrspace 1)
    ; GFX10-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](s96)
    %0:vgpr(p1) = COPY $vgpr0_vgpr1
    %1:vgpr(s96) = G_LOAD %0 :: (load (s96), align 4, addrspace 1)
    $vgpr0_vgpr1_vgpr2 = COPY %1

...

---

name: load_global_v6s16
legalized:       true
regBankSelected: true
tracksRegLiveness: true

body: |
  bb.0:
    liveins:  $vgpr0_vgpr1

    ; GFX7-LABEL: name: load_global_v6s16
    ; GFX7: liveins: $vgpr0_vgpr1
    ; GFX7-NEXT: {{  $}}
    ; GFX7-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX7-NEXT: [[LOAD:%[0-9]+]]:vreg_96(<6 x s16>) = G_LOAD [[COPY]](p1) :: (load (<6 x s16>), align 4, addrspace 1)
    ; GFX7-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](<6 x s16>)
    ;
    ; GFX7-FLAT-LABEL: name: load_global_v6s16
    ; GFX7-FLAT: liveins: $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: {{  $}}
    ; GFX7-FLAT-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX7-FLAT-NEXT: [[LOAD:%[0-9]+]]:vreg_96(<6 x s16>) = G_LOAD [[COPY]](p1) :: (load (<6 x s16>), align 4, addrspace 1)
    ; GFX7-FLAT-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](<6 x s16>)
    ;
    ; GFX8-LABEL: name: load_global_v6s16
    ; GFX8: liveins: $vgpr0_vgpr1
    ; GFX8-NEXT: {{  $}}
    ; GFX8-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX8-NEXT: [[LOAD:%[0-9]+]]:vreg_96(<6 x s16>) = G_LOAD [[COPY]](p1) :: (load (<6 x s16>), align 4, addrspace 1)
    ; GFX8-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](<6 x s16>)
    ;
    ; GFX9-LABEL: name: load_global_v6s16
    ; GFX9: liveins: $vgpr0_vgpr1
    ; GFX9-NEXT: {{  $}}
    ; GFX9-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX9-NEXT: [[LOAD:%[0-9]+]]:vreg_96(<6 x s16>) = G_LOAD [[COPY]](p1) :: (load (<6 x s16>), align 4, addrspace 1)
    ; GFX9-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](<6 x s16>)
    ;
    ; GFX10-LABEL: name: load_global_v6s16
    ; GFX10: liveins: $vgpr0_vgpr1
    ; GFX10-NEXT: {{  $}}
    ; GFX10-NEXT: [[COPY:%[0-9]+]]:vgpr(p1) = COPY $vgpr0_vgpr1
    ; GFX10-NEXT: [[LOAD:%[0-9]+]]:vreg_96(<6 x s16>) = G_LOAD [[COPY]](p1) :: (load (<6 x s16>), align 4, addrspace 1)
    ; GFX10-NEXT: $vgpr0_vgpr1_vgpr2 = COPY [[LOAD]](<6 x s16>)
    %0:vgpr(p1) = COPY $vgpr0_vgpr1
    %1:vgpr(<6 x s16>) = G_LOAD %0 :: (load (<6 x s16>), align 4, addrspace 1)
    $vgpr0_vgpr1_vgpr2 = COPY %1

...