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
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -instcombine -S | FileCheck %s
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"

; Verify that instcombine is able to fold identity shuffles.

define <8 x i32> @identity_test_vpermd(<8 x i32> %a0) {
; CHECK-LABEL: @identity_test_vpermd(
; CHECK-NEXT:    ret <8 x i32> %a0
;
  %a = tail call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>)
  ret <8 x i32> %a
}

define <8 x float> @identity_test_vpermps(<8 x float> %a0) {
; CHECK-LABEL: @identity_test_vpermps(
; CHECK-NEXT:    ret <8 x float> %a0
;
  %a = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>)
  ret <8 x float> %a
}

; Instcombine should be able to fold the following shuffle to a builtin shufflevector
; with a shuffle mask of all zeroes.

define <8 x i32> @zero_test_vpermd(<8 x i32> %a0) {
; CHECK-LABEL: @zero_test_vpermd(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> zeroinitializer
; CHECK-NEXT:    ret <8 x i32> [[TMP1]]
;
  %a = tail call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> zeroinitializer)
  ret <8 x i32> %a
}

define <8 x float> @zero_test_vpermps(<8 x float> %a0) {
; CHECK-LABEL: @zero_test_vpermps(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> %a0, <8 x float> undef, <8 x i32> zeroinitializer
; CHECK-NEXT:    ret <8 x float> [[TMP1]]
;
  %a = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> zeroinitializer)
  ret <8 x float> %a
}

; Verify that instcombine is able to fold constant shuffles.

define <8 x i32> @shuffle_test_vpermd(<8 x i32> %a0) {
; CHECK-LABEL: @shuffle_test_vpermd(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT:    ret <8 x i32> [[TMP1]]
;
  %a = tail call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>)
  ret <8 x i32> %a
}

define <8 x float> @shuffle_test_vpermps(<8 x float> %a0) {
; CHECK-LABEL: @shuffle_test_vpermps(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> %a0, <8 x float> undef, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT:    ret <8 x float> [[TMP1]]
;
  %a = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> <i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>)
  ret <8 x float> %a
}

; Verify that instcombine is able to fold constant shuffles with undef mask elements.

define <8 x i32> @undef_test_vpermd(<8 x i32> %a0) {
; CHECK-LABEL: @undef_test_vpermd(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 undef, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT:    ret <8 x i32> [[TMP1]]
;
  %a = tail call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> <i32 undef, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>)
  ret <8 x i32> %a
}

define <8 x float> @undef_test_vpermps(<8 x float> %a0) {
; CHECK-LABEL: @undef_test_vpermps(
; CHECK-NEXT:    [[TMP1:%.*]] = shufflevector <8 x float> %a0, <8 x float> undef, <8 x i32> <i32 undef, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT:    ret <8 x float> [[TMP1]]
;
  %a = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> <i32 undef, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0>)
  ret <8 x float> %a
}

; Verify simplify demanded elts.

define <8 x i32> @elts_test_vpermd(<8 x i32> %a0, i32 %a1) {
; CHECK-LABEL: @elts_test_vpermd(
; CHECK-NEXT:    ret <8 x i32> %a0
;
  %1 = insertelement <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>, i32 %a1, i32 0
  %2 = tail call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> %1)
  %3 = shufflevector <8 x i32> %2, <8 x i32> undef, <8 x i32> <i32 undef, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
  ret <8 x i32> %3
}

define <8 x float> @elts_test_vpermps(<8 x float> %a0, <8 x i32> %a1) {
; CHECK-LABEL: @elts_test_vpermps(
; CHECK-NEXT:    [[TMP1:%.*]] = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> %a1)
; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <8 x float> [[TMP1]], <8 x float> undef, <8 x i32> zeroinitializer
; CHECK-NEXT:    ret <8 x float> [[TMP2]]
;
  %1 = insertelement <8 x i32> %a1, i32 0, i32 7
  %2 = tail call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x i32> %1)
  %3 = shufflevector <8 x float> %2, <8 x float> undef, <8 x i32> zeroinitializer
  ret <8 x float> %3
}

declare <8 x i32> @llvm.x86.avx2.permd(<8 x i32>, <8 x i32>)
declare <8 x float> @llvm.x86.avx2.permps(<8 x float>, <8 x i32>)