reference, declarationdefinition
definition → references, declarations, derived classes, virtual overrides
reference to multiple definitions → definitions
unreferenced
    1
    2
    3
    4
    5
    6
    7
    8
    9
   10
   11
   12
   13
   14
   15
   16
   17
   18
   19
   20
   21
   22
   23
   24
   25
   26
   27
   28
   29
   30
   31
   32
   33
   34
   35
   36
   37
   38
   39
   40
   41
   42
   43
   44
   45
   46
   47
   48
   49
   50
   51
   52
   53
   54
   55
   56
   57
   58
   59
   60
   61
   62
   63
   64
   65
   66
   67
   68
   69
   70
   71
   72
   73
   74
   75
   76
   77
   78
   79
   80
   81
   82
   83
   84
   85
   86
   87
   88
   89
   90
   91
   92
   93
   94
   95
   96
   97
   98
   99
  100
  101
  102
  103
  104
  105
  106
  107
  108
  109
  110
  111
  112
  113
  114
  115
  116
  117
  118
  119
  120
; RUN: llc -march=amdgcn < %s | FileCheck -check-prefix=SI --check-prefix=FUNC %s
; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global < %s | FileCheck -check-prefix=SI --check-prefix=FUNC %s
; RUN: llc -march=r600 -mcpu=redwood < %s | FileCheck -check-prefix=EG --check-prefix=FUNC %s

declare float @llvm.trunc.f32(float) nounwind readnone
declare <2 x float> @llvm.trunc.v2f32(<2 x float>) nounwind readnone
declare <3 x float> @llvm.trunc.v3f32(<3 x float>) nounwind readnone
declare <4 x float> @llvm.trunc.v4f32(<4 x float>) nounwind readnone
declare <8 x float> @llvm.trunc.v8f32(<8 x float>) nounwind readnone
declare <16 x float> @llvm.trunc.v16f32(<16 x float>) nounwind readnone

; FUNC-LABEL: {{^}}ftrunc_f32:
; EG: TRUNC
; SI: v_trunc_f32_e32
define amdgpu_kernel void @ftrunc_f32(float addrspace(1)* %out, float %x) {
  %y = call float @llvm.trunc.f32(float %x) nounwind readnone
  store float %y, float addrspace(1)* %out
  ret void
}

; FUNC-LABEL: {{^}}ftrunc_v2f32:
; EG: TRUNC
; EG: TRUNC
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
define amdgpu_kernel void @ftrunc_v2f32(<2 x float> addrspace(1)* %out, <2 x float> %x) {
  %y = call <2 x float> @llvm.trunc.v2f32(<2 x float> %x) nounwind readnone
  store <2 x float> %y, <2 x float> addrspace(1)* %out
  ret void
}

; FIXME-FUNC-LABEL: {{^}}ftrunc_v3f32:
; FIXME-EG: TRUNC
; FIXME-EG: TRUNC
; FIXME-EG: TRUNC
; FIXME-SI: v_trunc_f32_e32
; FIXME-SI: v_trunc_f32_e32
; FIXME-SI: v_trunc_f32_e32
; define amdgpu_kernel void @ftrunc_v3f32(<3 x float> addrspace(1)* %out, <3 x float> %x) {
;   %y = call <3 x float> @llvm.trunc.v3f32(<3 x float> %x) nounwind readnone
;   store <3 x float> %y, <3 x float> addrspace(1)* %out
;   ret void
; }

; FUNC-LABEL: {{^}}ftrunc_v4f32:
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
define amdgpu_kernel void @ftrunc_v4f32(<4 x float> addrspace(1)* %out, <4 x float> %x) {
  %y = call <4 x float> @llvm.trunc.v4f32(<4 x float> %x) nounwind readnone
  store <4 x float> %y, <4 x float> addrspace(1)* %out
  ret void
}

; FUNC-LABEL: {{^}}ftrunc_v8f32:
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
define amdgpu_kernel void @ftrunc_v8f32(<8 x float> addrspace(1)* %out, <8 x float> %x) {
  %y = call <8 x float> @llvm.trunc.v8f32(<8 x float> %x) nounwind readnone
  store <8 x float> %y, <8 x float> addrspace(1)* %out
  ret void
}

; FUNC-LABEL: {{^}}ftrunc_v16f32:
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; EG: TRUNC
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
; SI: v_trunc_f32_e32
define amdgpu_kernel void @ftrunc_v16f32(<16 x float> addrspace(1)* %out, <16 x float> %x) {
  %y = call <16 x float> @llvm.trunc.v16f32(<16 x float> %x) nounwind readnone
  store <16 x float> %y, <16 x float> addrspace(1)* %out
  ret void
}