; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple powerpc64-unknown-linux-gnu -verify-machineinstrs < %s | FileCheck %s
; RUN: llc -mtriple powerpc64-ibm-aix-xcoff -verify-machineinstrs -vec-extabi < %s | FileCheck %s

define <16 x i8> @test_l_v16i8(<16 x i8>* %p) #0 {
; CHECK-LABEL: test_l_v16i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 15
; CHECK-NEXT:    lvsl 3, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    lvx 4, 0, 3
; CHECK-NEXT:    vperm 2, 4, 2, 3
; CHECK-NEXT:    blr
entry:
  %r = load <16 x i8>, <16 x i8>* %p, align 1
  ret <16 x i8> %r

}

define <32 x i8> @test_l_v32i8(<32 x i8>* %p) #0 {
; CHECK-LABEL: test_l_v32i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 31
; CHECK-NEXT:    lvsl 5, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lvx 4, 3, 4
; CHECK-NEXT:    lvx 0, 0, 3
; CHECK-NEXT:    vperm 3, 4, 2, 5
; CHECK-NEXT:    vperm 2, 0, 4, 5
; CHECK-NEXT:    blr
entry:
  %r = load <32 x i8>, <32 x i8>* %p, align 1
  ret <32 x i8> %r

}

define <8 x i16> @test_l_v8i16(<8 x i16>* %p) #0 {
; CHECK-LABEL: test_l_v8i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 15
; CHECK-NEXT:    lvsl 3, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    lvx 4, 0, 3
; CHECK-NEXT:    vperm 2, 4, 2, 3
; CHECK-NEXT:    blr
entry:
  %r = load <8 x i16>, <8 x i16>* %p, align 2
  ret <8 x i16> %r

}

define <16 x i16> @test_l_v16i16(<16 x i16>* %p) #0 {
; CHECK-LABEL: test_l_v16i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 31
; CHECK-NEXT:    lvsl 5, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lvx 4, 3, 4
; CHECK-NEXT:    lvx 0, 0, 3
; CHECK-NEXT:    vperm 3, 4, 2, 5
; CHECK-NEXT:    vperm 2, 0, 4, 5
; CHECK-NEXT:    blr
entry:
  %r = load <16 x i16>, <16 x i16>* %p, align 2
  ret <16 x i16> %r

}

define <4 x i32> @test_l_v4i32(<4 x i32>* %p) #0 {
; CHECK-LABEL: test_l_v4i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 15
; CHECK-NEXT:    lvsl 3, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    lvx 4, 0, 3
; CHECK-NEXT:    vperm 2, 4, 2, 3
; CHECK-NEXT:    blr
entry:
  %r = load <4 x i32>, <4 x i32>* %p, align 4
  ret <4 x i32> %r

}

define <8 x i32> @test_l_v8i32(<8 x i32>* %p) #0 {
; CHECK-LABEL: test_l_v8i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 31
; CHECK-NEXT:    lvsl 5, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lvx 4, 3, 4
; CHECK-NEXT:    lvx 0, 0, 3
; CHECK-NEXT:    vperm 3, 4, 2, 5
; CHECK-NEXT:    vperm 2, 0, 4, 5
; CHECK-NEXT:    blr
entry:
  %r = load <8 x i32>, <8 x i32>* %p, align 4
  ret <8 x i32> %r

}

define <2 x i64> @test_l_v2i64(<2 x i64>* %p) #0 {
; CHECK-LABEL: test_l_v2i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <2 x i64>, <2 x i64>* %p, align 8
  ret <2 x i64> %r

}

define <4 x i64> @test_l_v4i64(<4 x i64>* %p) #0 {
; CHECK-LABEL: test_l_v4i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    lxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <4 x i64>, <4 x i64>* %p, align 8
  ret <4 x i64> %r

}

define <4 x float> @test_l_v4float(<4 x float>* %p) #0 {
; CHECK-LABEL: test_l_v4float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 15
; CHECK-NEXT:    lvsl 3, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    lvx 4, 0, 3
; CHECK-NEXT:    vperm 2, 4, 2, 3
; CHECK-NEXT:    blr
entry:
  %r = load <4 x float>, <4 x float>* %p, align 4
  ret <4 x float> %r

}

define <8 x float> @test_l_v8float(<8 x float>* %p) #0 {
; CHECK-LABEL: test_l_v8float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 31
; CHECK-NEXT:    lvsl 5, 0, 3
; CHECK-NEXT:    lvx 2, 3, 4
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lvx 4, 3, 4
; CHECK-NEXT:    lvx 0, 0, 3
; CHECK-NEXT:    vperm 3, 4, 2, 5
; CHECK-NEXT:    vperm 2, 0, 4, 5
; CHECK-NEXT:    blr
entry:
  %r = load <8 x float>, <8 x float>* %p, align 4
  ret <8 x float> %r

}

define <2 x double> @test_l_v2double(<2 x double>* %p) #0 {
; CHECK-LABEL: test_l_v2double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <2 x double>, <2 x double>* %p, align 8
  ret <2 x double> %r

}

define <4 x double> @test_l_v4double(<4 x double>* %p) #0 {
; CHECK-LABEL: test_l_v4double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    lxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <4 x double>, <4 x double>* %p, align 8
  ret <4 x double> %r

}

define <16 x i8> @test_l_p8v16i8(<16 x i8>* %p) #2 {
; CHECK-LABEL: test_l_p8v16i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <16 x i8>, <16 x i8>* %p, align 1
  ret <16 x i8> %r

}

define <32 x i8> @test_l_p8v32i8(<32 x i8>* %p) #2 {
; CHECK-LABEL: test_l_p8v32i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    lxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <32 x i8>, <32 x i8>* %p, align 1
  ret <32 x i8> %r

}

define <8 x i16> @test_l_p8v8i16(<8 x i16>* %p) #2 {
; CHECK-LABEL: test_l_p8v8i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <8 x i16>, <8 x i16>* %p, align 2
  ret <8 x i16> %r

}

define <16 x i16> @test_l_p8v16i16(<16 x i16>* %p) #2 {
; CHECK-LABEL: test_l_p8v16i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    lxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <16 x i16>, <16 x i16>* %p, align 2
  ret <16 x i16> %r

}

define <4 x i32> @test_l_p8v4i32(<4 x i32>* %p) #2 {
; CHECK-LABEL: test_l_p8v4i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <4 x i32>, <4 x i32>* %p, align 4
  ret <4 x i32> %r

}

define <8 x i32> @test_l_p8v8i32(<8 x i32>* %p) #2 {
; CHECK-LABEL: test_l_p8v8i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    lxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <8 x i32>, <8 x i32>* %p, align 4
  ret <8 x i32> %r

}

define <2 x i64> @test_l_p8v2i64(<2 x i64>* %p) #2 {
; CHECK-LABEL: test_l_p8v2i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <2 x i64>, <2 x i64>* %p, align 8
  ret <2 x i64> %r

}

define <4 x i64> @test_l_p8v4i64(<4 x i64>* %p) #2 {
; CHECK-LABEL: test_l_p8v4i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    lxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <4 x i64>, <4 x i64>* %p, align 8
  ret <4 x i64> %r

}

define <4 x float> @test_l_p8v4float(<4 x float>* %p) #2 {
; CHECK-LABEL: test_l_p8v4float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <4 x float>, <4 x float>* %p, align 4
  ret <4 x float> %r

}

define <8 x float> @test_l_p8v8float(<8 x float>* %p) #2 {
; CHECK-LABEL: test_l_p8v8float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvw4x 34, 0, 3
; CHECK-NEXT:    lxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <8 x float>, <8 x float>* %p, align 4
  ret <8 x float> %r

}

define <2 x double> @test_l_p8v2double(<2 x double>* %p) #2 {
; CHECK-LABEL: test_l_p8v2double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  %r = load <2 x double>, <2 x double>* %p, align 8
  ret <2 x double> %r

}

define <4 x double> @test_l_p8v4double(<4 x double>* %p) #2 {
; CHECK-LABEL: test_l_p8v4double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    lxvd2x 34, 0, 3
; CHECK-NEXT:    lxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  %r = load <4 x double>, <4 x double>* %p, align 8
  ret <4 x double> %r

}

define void @test_s_v16i8(<16 x i8>* %p, <16 x i8> %v) #0 {
; CHECK-LABEL: test_s_v16i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <16 x i8> %v, <16 x i8>* %p, align 1
  ret void

}

define void @test_s_v32i8(<32 x i8>* %p, <32 x i8> %v) #0 {
; CHECK-LABEL: test_s_v32i8:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    stxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <32 x i8> %v, <32 x i8>* %p, align 1
  ret void

}

define void @test_s_v8i16(<8 x i16>* %p, <8 x i16> %v) #0 {
; CHECK-LABEL: test_s_v8i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <8 x i16> %v, <8 x i16>* %p, align 2
  ret void

}

define void @test_s_v16i16(<16 x i16>* %p, <16 x i16> %v) #0 {
; CHECK-LABEL: test_s_v16i16:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    stxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <16 x i16> %v, <16 x i16>* %p, align 2
  ret void

}

define void @test_s_v4i32(<4 x i32>* %p, <4 x i32> %v) #0 {
; CHECK-LABEL: test_s_v4i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <4 x i32> %v, <4 x i32>* %p, align 4
  ret void

}

define void @test_s_v8i32(<8 x i32>* %p, <8 x i32> %v) #0 {
; CHECK-LABEL: test_s_v8i32:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    stxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <8 x i32> %v, <8 x i32>* %p, align 4
  ret void

}

define void @test_s_v2i64(<2 x i64>* %p, <2 x i64> %v) #0 {
; CHECK-LABEL: test_s_v2i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <2 x i64> %v, <2 x i64>* %p, align 8
  ret void

}

define void @test_s_v4i64(<4 x i64>* %p, <4 x i64> %v) #0 {
; CHECK-LABEL: test_s_v4i64:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvd2x 34, 0, 3
; CHECK-NEXT:    stxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <4 x i64> %v, <4 x i64>* %p, align 8
  ret void

}

define void @test_s_v4float(<4 x float>* %p, <4 x float> %v) #0 {
; CHECK-LABEL: test_s_v4float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <4 x float> %v, <4 x float>* %p, align 4
  ret void

}

define void @test_s_v8float(<8 x float>* %p, <8 x float> %v) #0 {
; CHECK-LABEL: test_s_v8float:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvw4x 34, 0, 3
; CHECK-NEXT:    stxvw4x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <8 x float> %v, <8 x float>* %p, align 4
  ret void

}

define void @test_s_v2double(<2 x double>* %p, <2 x double> %v) #0 {
; CHECK-LABEL: test_s_v2double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    stxvd2x 34, 0, 3
; CHECK-NEXT:    blr
entry:
  store <2 x double> %v, <2 x double>* %p, align 8
  ret void

}

define void @test_s_v4double(<4 x double>* %p, <4 x double> %v) #0 {
; CHECK-LABEL: test_s_v4double:
; CHECK:       # %bb.0: # %entry
; CHECK-NEXT:    li 4, 16
; CHECK-NEXT:    stxvd2x 34, 0, 3
; CHECK-NEXT:    stxvd2x 35, 3, 4
; CHECK-NEXT:    blr
entry:
  store <4 x double> %v, <4 x double>* %p, align 8
  ret void

}

attributes #0 = { nounwind "target-cpu"="pwr7" }
attributes #2 = { nounwind "target-cpu"="pwr8" }