llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-ctlz.mir

# 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

---
name: ctlz_s32_s32

body: |
  bb.0:
    liveins: $vgpr0
    ; CHECK-LABEL: name: ctlz_s32_s32
    ; CHECK: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[COPY]](s32)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: $vgpr0 = COPY [[UMIN]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s32) = G_CTLZ %0
    $vgpr0 = COPY %1
...

---
name: ctlz_s32_s64

body: |
  bb.0:
    liveins: $vgpr0_vgpr1
    ; CHECK-LABEL: name: ctlz_s32_s64
    ; CHECK: [[COPY:%[0-9]+]]:_(s64) = COPY $vgpr0_vgpr1
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[COPY]](s64)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 64
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: $vgpr0 = COPY [[UMIN]](s32)
    %0:_(s64) = COPY $vgpr0_vgpr1
    %1:_(s32) = G_CTLZ %0
    $vgpr0 = COPY %1
...

---
name: ctlz_s64_s64

body: |
  bb.0:
    liveins: $vgpr0_vgpr1
    ; CHECK-LABEL: name: ctlz_s64_s64
    ; CHECK: [[COPY:%[0-9]+]]:_(s64) = COPY $vgpr0_vgpr1
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[COPY]](s64)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 64
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: [[ZEXT:%[0-9]+]]:_(s64) = G_ZEXT [[UMIN]](s32)
    ; CHECK: $vgpr0_vgpr1 = COPY [[ZEXT]](s64)
    %0:_(s64) = COPY $vgpr0_vgpr1
    %1:_(s64) = G_CTLZ %0
    $vgpr0_vgpr1 = COPY %1
...

---
name: ctlz_s16_s32

body: |
  bb.0:
    liveins: $vgpr0
    ; CHECK-LABEL: name: ctlz_s16_s32
    ; CHECK: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[COPY]](s32)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
    ; CHECK: [[AND:%[0-9]+]]:_(s32) = G_AND [[UMIN]], [[C1]]
    ; CHECK: $vgpr0 = COPY [[AND]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s16) = G_CTLZ %0
    %2:_(s32) = G_ZEXT %1
    $vgpr0 = COPY %2
...

---
name: ctlz_s16_s16

body: |
  bb.0:
    liveins: $vgpr0
    ; CHECK-LABEL: name: ctlz_s16_s16
    ; CHECK: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
    ; CHECK: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C]]
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[AND]](s32)
    ; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C1]]
    ; CHECK: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; CHECK: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[UMIN]], [[C2]]
    ; CHECK: [[COPY1:%[0-9]+]]:_(s32) = COPY [[SUB]](s32)
    ; CHECK: [[AND1:%[0-9]+]]:_(s32) = G_AND [[COPY1]], [[C]]
    ; CHECK: $vgpr0 = COPY [[AND1]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s16) = G_TRUNC %0
    %2:_(s16) = G_CTLZ %1
    %3:_(s32) = G_ZEXT %2
    $vgpr0 = COPY %3
...

---
name: ctlz_v2s32_v2s32

body: |
  bb.0:
    liveins: $vgpr0_vgpr1
    ; CHECK-LABEL: name: ctlz_v2s32_v2s32
    ; CHECK: [[COPY:%[0-9]+]]:_(<2 x s32>) = COPY $vgpr0_vgpr1
    ; CHECK: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[COPY]](<2 x s32>)
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[UV]](s32)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: [[AMDGPU_FFBH_U32_1:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[UV1]](s32)
    ; CHECK: [[UMIN1:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_1]], [[C]]
    ; CHECK: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[UMIN]](s32), [[UMIN1]](s32)
    ; CHECK: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    %0:_(<2 x s32>) = COPY $vgpr0_vgpr1
    %1:_(<2 x s32>) = G_CTLZ %0
    $vgpr0_vgpr1 = COPY %1
...

---
name: ctlz_v2s32_v2s64

body: |
  bb.0:
    liveins: $vgpr0_vgpr1_vgpr2_vgpr3
    ; CHECK-LABEL: name: ctlz_v2s32_v2s64
    ; CHECK: [[COPY:%[0-9]+]]:_(<2 x s64>) = COPY $vgpr0_vgpr1_vgpr2_vgpr3
    ; CHECK: [[UV:%[0-9]+]]:_(s64), [[UV1:%[0-9]+]]:_(s64) = G_UNMERGE_VALUES [[COPY]](<2 x s64>)
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[UV]](s64)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 64
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C]]
    ; CHECK: [[AMDGPU_FFBH_U32_1:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[UV1]](s64)
    ; CHECK: [[UMIN1:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_1]], [[C]]
    ; CHECK: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[UMIN]](s32), [[UMIN1]](s32)
    ; CHECK: $vgpr0_vgpr1 = COPY [[BUILD_VECTOR]](<2 x s32>)
    %0:_(<2 x s64>) = COPY $vgpr0_vgpr1_vgpr2_vgpr3
    %1:_(<2 x s32>) = G_CTLZ %0
    $vgpr0_vgpr1 = COPY %1
...

---
name: ctlz_v2s16_v2s16

body: |
  bb.0:
    liveins: $vgpr0
    ; CHECK-LABEL: name: ctlz_v2s16_v2s16
    ; CHECK: [[COPY:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr0
    ; CHECK: [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY]](<2 x s16>)
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
    ; CHECK: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
    ; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
    ; CHECK: [[AND:%[0-9]+]]:_(s32) = G_AND [[BITCAST]], [[C1]]
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[AND]](s32)
    ; CHECK: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C2]]
    ; CHECK: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[UMIN]], [[C]]
    ; CHECK: [[COPY1:%[0-9]+]]:_(s32) = COPY [[SUB]](s32)
    ; CHECK: [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
    ; CHECK: [[AMDGPU_FFBH_U32_1:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[AND1]](s32)
    ; CHECK: [[UMIN1:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_1]], [[C2]]
    ; CHECK: [[SUB1:%[0-9]+]]:_(s32) = G_SUB [[UMIN1]], [[C]]
    ; CHECK: [[COPY2:%[0-9]+]]:_(s32) = COPY [[SUB1]](s32)
    ; CHECK: [[AND2:%[0-9]+]]:_(s32) = G_AND [[COPY1]], [[C1]]
    ; CHECK: [[AND3:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[C1]]
    ; CHECK: [[SHL:%[0-9]+]]:_(s32) = G_SHL [[AND3]], [[C]](s32)
    ; CHECK: [[OR:%[0-9]+]]:_(s32) = G_OR [[AND2]], [[SHL]]
    ; CHECK: [[BITCAST1:%[0-9]+]]:_(<2 x s16>) = G_BITCAST [[OR]](s32)
    ; CHECK: $vgpr0 = COPY [[BITCAST1]](<2 x s16>)
    %0:_(<2 x s16>) = COPY $vgpr0
    %1:_(<2 x s16>) = G_CTLZ %0
    $vgpr0 = COPY %1
...

---
name: ctlz_s7_s7

body: |
  bb.0:
    liveins: $vgpr0

    ; CHECK-LABEL: name: ctlz_s7_s7
    ; CHECK: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
    ; CHECK: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 127
    ; CHECK: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY]], [[C]]
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[AND]](s32)
    ; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 32
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C1]]
    ; CHECK: [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 25
    ; CHECK: [[SUB:%[0-9]+]]:_(s32) = G_SUB [[UMIN]], [[C2]]
    ; CHECK: [[COPY1:%[0-9]+]]:_(s32) = COPY [[SUB]](s32)
    ; CHECK: [[AND1:%[0-9]+]]:_(s32) = G_AND [[COPY1]], [[C]]
    ; CHECK: $vgpr0 = COPY [[AND1]](s32)
    %0:_(s32) = COPY $vgpr0
    %1:_(s7) = G_TRUNC %0
    %2:_(s7) = G_CTLZ %1
    %3:_(s32) = G_ZEXT %2
    $vgpr0 = COPY %3
...

---
name: ctlz_s33_s33

body: |
  bb.0:
    liveins: $vgpr0_vgpr1

    ; CHECK-LABEL: name: ctlz_s33_s33
    ; CHECK: [[COPY:%[0-9]+]]:_(s64) = COPY $vgpr0_vgpr1
    ; CHECK: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 8589934591
    ; CHECK: [[AND:%[0-9]+]]:_(s64) = G_AND [[COPY]], [[C]]
    ; CHECK: [[AMDGPU_FFBH_U32_:%[0-9]+]]:_(s32) = G_AMDGPU_FFBH_U32 [[AND]](s64)
    ; CHECK: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 64
    ; CHECK: [[UMIN:%[0-9]+]]:_(s32) = G_UMIN [[AMDGPU_FFBH_U32_]], [[C1]]
    ; CHECK: [[ZEXT:%[0-9]+]]:_(s64) = G_ZEXT [[UMIN]](s32)
    ; CHECK: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 31
    ; CHECK: [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[ZEXT]](s64)
    ; CHECK: [[UV2:%[0-9]+]]:_(s32), [[UV3:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[C2]](s64)
    ; CHECK: [[USUBO:%[0-9]+]]:_(s32), [[USUBO1:%[0-9]+]]:_(s1) = G_USUBO [[UV]], [[UV2]]
    ; CHECK: [[ZEXT1:%[0-9]+]]:_(s64) = G_ZEXT [[USUBO]](s32)
    ; CHECK: $vgpr0_vgpr1 = COPY [[ZEXT1]](s64)
    %0:_(s64) = COPY $vgpr0_vgpr1
    %1:_(s33) = G_TRUNC %0
    %2:_(s33) = G_CTLZ %1
    %3:_(s64) = G_ANYEXT %2
    $vgpr0_vgpr1 = COPY %3
...

# ---
# name: ctlz_v2s7_v2s7

# body: |
#   bb.0:
#     liveins: $vgpr0
#     %0:_(<2 x s32>) = COPY $vgpr0_vgpr1
#     %1:_(<2 x s7>) = G_TRUNC %0
#     %2:_(<2 x s7>) = G_CTLZ %1
#     %3:_(<2 x s32>) = G_ANYEXT %2
#     $vgpr0_vgpr1 = COPY %3
# ...